Safety evaluation of a recombinant plasmin derivative lacking kringles 2-5 and rt-PA in a rat model of transient ischemic stroke.

BACKGROUND: Tissue type plasminogen activator is the only approved thrombolytic agent for the treatment of ischemic stroke. However, it carries the disadvantage of a 10-fold increase in symptomatic and asymptomatic intracranial hemorrhage. A safer thrombolytic agent may improve patient prognosis and increase patient participation in thrombolytic treatment. A novel direct-acting thrombolytic agent, Δ(K2-K5) plasmin, promising an improved safety profile was examined for safety in the snare ligature model of stroke in the rat. METHODS: Male spontaneously hypertensive rats were subjected to 6 hours middle cerebral artery occlusion followed by 18 hours reflow. Beginning 1 minute before reflow, they were dosed with saline, vehicle, Δ(K2-K5) plasmin (0.15, 0.5, 1.5, and 5 mg/kg) or recombinant tissue-type plasminogen activator (10 and 30 mg/kg) by local intra-arterial infusion lasting 10 to 60 minutes. The rats were assessed for bleeding score, infarct volume, modified Bederson score and general behavioral score. In a parallel study, temporal progression of infarct volume was determined. In an in vitro study, whole blood clots from humans, canines and rats were exposed to Δ(K2-K5). Clot lysis was monitored by absorbance at 280 nm. RESULTS: The main focus of this study was intracranial hemorrhage safety. Δ(K2-K5) plasmin treatment at the highest dose caused no more intracranial hemorrhage than the lowest dose of recombinant tissue type plasminogen activator, but showed at least a 5-fold superior safety margin. Secondary results include: temporal infarct volume progression shows that the greatest expansion of infarct volume occurs within 2-3 hours of middle cerebral artery occlusion in the spontaneously hypertensive rat. A spike in infarct volume was observed at 6 hours ischemia with reflow. Δ(K2-K5) plasmin tended to reduce infarct volume and improve behavior compared to controls. In vitro data suggests that Δ(K2-K5) plasmin is equally effective at lysing clots from humans, canines and rats. CONCLUSIONS: The superior intracranial hemorrhage safety profile of the direct-acting thrombolytic Δ(K2-K5) plasmin compared with recombinant tissue type plasminogen activator makes this agent a good candidate for clinical evaluation in the treatment of acute ischemic stroke.

Intra-arterial administration of recombinant tissue-type plasminogen activator (rt-PA) causes more intracranial bleeding than does intravenous rt-PA in a transient rat middle cerebral artery occlusion model.

BACKGROUND: Intra-arterial (IA) administration of rt-PA for ischemic stroke has the potential for greater thrombolytic efficacy, especially for a large thrombus in the M1 or M2 segment of the middle cerebral artery (MCA). Intracranial hemorrhage (ICH) is a concern with IA or intravenous (IV) administration especially as the therapeutic window is extended. However, because IA administration delivers a higher local concentration of agent, the incidence and severity of ICH may be greater than with similar doses IV. We investigated the safety of rt-PA administration by IA compared to IV infusion following 6 hours of MCA occlusion (MCAo) with reflow in the spontaneously hypertensive rat (SHR). METHODS: Male SHRs were subjected to 6 hours MCAo with 18 hours reflow using a snare ligature model. They were treated with IA saline, IA rt-PA (1, 5, 10, 30 mg/kg), or IV rt-PA (10 and 30 mg/kg) by a 10 to 60 minute infusion beginning approximately 1 minute before reflow. The rats were recovered for 24 hours after MCAo onset at which time Bleeding Score, infarct volume, and Modified Bederson Score were measured. RESULTS: Greater hemorrhagic transformation occurred with 10 and 30 mg/kg rt-PA administered IA than IV. The IV 10 mg/kg rt-PA dosage induced significantly less bleeding than did the 1 or 5 mg/kg IA groups. No significant increase in infarct volume was observed after IA or IV treatment. Rats treated with 30 mg/kg rt-PA by either the IA or IV route had greater neurological dysfunction compared to all other groups. CONCLUSIONS: Administration of rt-PA by the IA route following 6 hours of MCAo results in greater ICH and worse functional recovery than comparable dosages IV. Significantly greater bleeding was observed when the IA dose was a tenth of the IV dose. The increased bleeding did not translate in larger infarct volumes.

Haemostatic safety of a unique recombinant plasmin molecule lacking kringles 2-5.

We previously demonstrated a significant margin of haemostatic safety for full-length plasmin in comparison with tissue plasminogen activator (t-PA). We now report studies that compare haemostatic safety of full-length plasmin with a novel recombinant plasmin derivative, (Δ K2-5) plasmin, consisting of kringle 1 linked to the serine protease domain of plasmin. Agent was administered intravenously in a randomised, blinded manner in a rabbit model of fibrinolytic haemorrhage. A dose-related decrease in α2-antiplasmin, factor VIII, and fibrinogen followed administration of 1.8, 2.7, 3.7 and 4.6 mg/kg of (Δ K2-5) plasmin, with nadir fibrinogen concentrations of 65%, 40%, 30%, and 0% of initial levels, respectively. Mean primary bleeding time was undisturbed at 1.8 mg/kg (2.2 ± 0.7 minutes), minimally prolonged at 2.7 or 3.7 mg/kg (5 ± 2.9 and 4.4 ± 2.2 minutes), and prolonged at the purposefully toxic 4.6 mg/kg dose (12.8 ± 18.8 minutes). Equimolar amounts of (Δ K2-5) plasmin and full-length plasmin had equal in vitro clot lysis efficacy, but in the bleeding model, (Δ K2-5) plasmin showed better haemostatic competency than full-length plasmin. This safety advantage may be explained by higher residual amounts of plasma fibrinogen in animals given (Δ K2-5) plasmin rather than full-length plasmin. We demonstrate that a unique recombinant plasmin mutant, (Δ K2-5) plasmin, possesses an advantage in hemostatic safety over an equimolar amount of full-length plasmin.

Simplified recombinant plasmin: production and functional comparison of a novel thrombolytic molecule with plasma-derived plasmin.

A simplified and fully functional deletion mutant of plasminogen was created in which the middle portion of the molecule was removed, resulting in kringle 1 attachment to the serine protease domain. This recombinant plasminogen deletion mutant, Delta(K2-K5)Pg, was produced in the form of inclusion bodies at the yield of up to 200 mg/l in an Escherichia coli T7 expression system. Following protein refolding and purification on lysine-Sepharose, the conversion of the recombinant molecule Delta(K2-K5)Pg to the active enzyme mutant Delta(K2-K5)Pm by plasminogen activators was evaluated, and functional characteristics of the simplified plasmin were studied. Properties of Delta(K2-K5)Pg were similar to native, human plasma-derived plasminogen. Delta(K2-K5)Pg effectively bound epsilon-aminocaproic acid (K(d)=11.3+/-2.3 microM) and fibrin (C(50) approximately 0.3 microM). The plasminogen activators streptokinase, urokinase, and tissue plasminogen activator effectively converted the recombinant zymogen Delta(K2-K5)Pg to the active recombinant enzyme, Delta(K2-K5)Pm. Additionally, Delta[K2-K5]Pm was rapidly inhibited by alpha(2)-antiplasmin (1.1+/-0.1 x 10(7) M(-1) s(-1)) and alpha(2)-macroglobulin (7.6+/-0.6 x 10(5) M(-1) s(-1)). In an in-vitro model, Delta(K2-K5)Pm demonstrated fibrinolytic potency comparable to human plasma-derived plasmin. Because of their unique biochemistry, including fibrin-binding properties and rapid inhibition by alpha(2)-antiplasmin, both native plasmin and a simplified deletion mutant of plasmin are potentially safe and effective direct thrombolytic agents for various thrombotic conditions. Further studies evaluating the in-vivo pharmacologic safety and clinical efficacy of this simplified plasmin (i.e. Delta[K2-K5]Pm) are warranted.

Clearance of prions during plasma protein manufacture.

Protein products isolated from human plasma are an important class of therapeutics that are used to treat patients afflicted with hereditary deficiencies, trauma, and severe infections. Because of the human origin of the starting material for the production of these biological products, there is a risk of transmitting infectious agents, including viruses and the infectious agents that cause transmissible spongiform encephalopathies (TSEs). The agent that is thought to cause TSEs is a disease-associated, misfolded form of the prion protein or PrP(Sc). Unlike viruses, there are no donor screening tests for TSEs available, and PrP(Sc) is resistant to traditional viral inactivation methods. Therefore, manufacturers of plasma products are faced with special challenges to ensure product safety with respect to TSEs. Fortunately, a growing body of evidence supports the capacity of manufacturing processes to remove infectious prions from the product stream during the purification of plasma products. This can be attributed in part to the unusual physicochemical nature of PrP(Sc), which is distinct from that of soluble therapeutic proteins. Although there is no reported TSE transmission through the use of plasma products to date, many unknowns remain to be addressed through long-term epidemiologic monitoring and further experimental studies.

An improved Western blot assay to assess the clearance of prion protein from plasma-derived therapeutic proteins.

Specific detection of the pathogenic prion protein, PrP(Sc), is essential for determining the prion clearance capacity of purification processes for therapeutic proteins. Use of a previously described indirect (two-antibody) Western blot assay sometimes resulted in the appearance of non-specific protein bands that interfered with the detection of small amounts of PrP(Sc)-specific signal, limiting the amount of clearance that could be determined for steps so affected. It is shown that these non-specific signals are due to the interaction between immunoglobulin fragments in the sample and the secondary antibody used in the assay. To circumvent this problem, a direct Western blot assay using a prion-specific primary antibody conjugated to the reporter enzyme alkaline phosphatase was developed. Application of the direct Western blot assay resulted in a significant reduction of non-specific signal while retaining the detection sensitivity for PrP(Sc)-specific signal. Therefore, the direct Western blot assay format is an improved tool for determining prion clearance capacity, particularly for immunoglobulin-rich samples.

Partitioning of human and sheep forms of the pathogenic prion protein during the purification of therapeutic proteins from human plasma.

BACKGROUND: Therapeutic proteins derived from human plasma and other biologic sources have demonstrated an excellent safety record relative to the potential threat of transmissible spongiform encephalopathy (TSE) transmission. Previously, hamster-adapted scrapie was used as a model agent to assess TSE clearance in purification steps leading to the isolation of biopharmaceutical proteins. The current study investigated the validity of hamster scrapie as a model for human TSE clearance studies. The partitioning of the pathogenic forms of the prion protein associated with human variant CJD (PrP(vCJD)), human sporadic CJD (PrP(sCJD)) and Gerstmann-Sträussler-Scheinker (PrP(GSS)) syndrome was compared to the partitioning of hamster scrapie (PrP(Sc)) in three plasma protein purification steps. Sheep scrapie (PrP(Sc)) was similarly evaluated. STUDY DESIGN AND METHODS: The starting materials for three plasma protein purification steps, cryoseparation, 3 percent PEG separation, and 11.5 percent PEG separation, were spiked with brain homogenates containing human PrP(vCJD), human PrP(sCJD), human PrP(GSS), sheep PrP(Sc), and hamster 263K PrP(Sc). The partitioning of the pathogenic form of the PrP was analyzed. RESULTS: Clearance of the pathogenic form of the PrP was measured relative to the effluent fraction. Regardless of the source of the pathogenic prion, clearance was similar to hamster PrP(Sc). A nominal amount of clearance (approx., 1 log), an intermediate amount of clearance (approx., 2 log), and a substantial amount of clearance (> or = 3 log) were observed for the cryoseparation, 3 percent PEG separation, and 11.5 percent PEG separation steps, respectively. In the latter step, no PrP was detected in the effluents. CONCLUSIONS: These data demonstrate that human prions, including vCJD prions, can be removed during the purification of human therapeutic proteins and indicate that partitioning of human prions is similar to that observed in the hamster scrapie model.

Solvent-dependent precipitation of prion protein.

The misfolded isoform of the prion protein (PrP(Sc)) possesses many unusual physiochemical properties. Previously, we and others reported on the differential partitioning of PrP(Sc) from plasma derived therapeutic proteins during their purification processes. To understand the driving force behind these partitioning differences, we investigated the effects of various solvent conditions on the precipitation of PrP(Sc). In a physiological buffer, PrP(Sc) remained in the supernatant after low speed centrifugation. At pH 5, PrP(Sc) precipitation was nearly complete regardless of the salt content. PrP(Sc) could also be precipitated at pH 8 by adding ethanol, but this precipitation was salt dependent. Based on these observations, an empirical mathematical model was constructed in which the PrP(Sc) precipitation trends were fully described as a function of solvent pH, salt, and ethanol concentration. This model consistently predicted PrP(Sc) partitioning during cold ethanol precipitation steps used in plasma protein purification processes, as shown by experimentally determined distributions of PrP(Sc) and transmissible spongiform encephalopathy (TSE) infectivity. These results indicate that pH, salt, and ethanol content are the major solvent factors determining the precipitation of the infectious PrP(Sc) in these processes and may provide a useful tool for assessing the differential partitioning of PrP(Sc) in a given solvent environment.