Expression of active chimeric-tissue plasminogen activator in tobacco hairy roots, identification of a DNA aptamer and purification by aptamer functionalized-MWCNTs chromatography

Tissue-type plasminogen activator (tPA) is a serine protease that plays a crucial role in the fibrinolytic system. We increased the activity of tPA by splicing the active site of dodder-cuscutain gene to human tPA. The chimeric cDNA of tPA was constructed by Splicing by Overlap Extension Polymerase Chain Reaction (SOEing-PCR) method and transferred to the hairy roots of tobacco using different strains of Agrobacterium rhizogenes. Chimeric-tPA was purified by lysine-sepharose chromatography and specific aptamers were designed using SELEX method. Multi wall carbon nanotubes were functionalized with selected aptamers, packed in a column, and used for purification. The results demonstrated that selected aptamer having KD values of 0.320 nM and IC50 of 28.9 nM possessed good affinity to tPA, and the chimeric-tPA was properly purified by aptamer-chromatography. Hairy roots expressing chimeric-tPA and normal-tPA produced 900 and 450 ngmg-1 of total protein, respectively. The activities of chimeric-tPA and normal-tPA were 90 and 60 IUml-1, respectively. Compared to the normal-tPA, chimeric-tPA showed more activity.

Purification of recombinant tissue plasminogen activator (rtPA) protein from transplastomic tobacco plants

Plants are low cost platforms for the production of recombinant proteins, but their complexity renders the purification of plant recombinant proteins more difficult than proteins expressed in yeast or bacteria. Plastid transformation enables high-level expression of foreign genes and the accumulation of recombinant proteins in plastid organelles. Histidine (His) tags are widely used for affinity purification of recombinant proteins in a nickel column. The human tissue-type plasminogen activator (tPA) is one of the most important pharmaceutical recombinant proteins involved in the breakdown of blood clots in different parts of the body. The truncated form of the tissue plasminogen activator (K2S) has a longer plasma half-life, better diffusion into the clot, and higher fibrinolytic activity. In a construct designed to insert the K2S gene in the tobacco chloroplast, the sequence of six histidines and a factor Xa protease site was fused to the C-terminus of the K2S protein. The presence and amount of tPA recombinant protein in transplastomic tobacco plants was estimated by ELISA analysis using a specific antibody. The protein was purified from total soluble protein, insoluble protein aggregates and the protein was extracted from the isolated chloroplast using nickel resin and a chromatography column. After digestion of the purified protein with factor Xa, the presence of the purified tPA protein was confirmed by western blot analysis.

Expression analysis and purification of human recombinant tissue type plasminogen activator (rt-PA) from transgenic tobacco plants

Recombinant tissue-type plasminogen activator (rt-PA) has been produced in different hosts. In this research, transgenic tobacco was selected for production of human rt-PA. Transgenic plants were analyzed by polymerase chain reaction (PCR) and reverse-transcription (RT)-PCR. The protein was extracted by Lysine Sepharose chromatography column and was further purified by HiTrap desalting column. The function of eluted protein was analyzed on zymography gel. The results showed that the 1.7-kb cDNA of tissue-type plasminogen activator (t-PA) (as well as a shortened 650-bp transcript of t-PA) has been expressed in transgenic plants. The anticipated 63-kD protein band and an additional 53-kD protein were observed in transgenic plants. Finally, zymography assay revealed that the purified rt-PA has anticipated appropriate activity comparable to a positive control drug (Alteplase). On the whole, we can say that transgenic tobacco is a good alternative host for production of t-PA.

Novel biomimetic affinity ligands for human tissue plasminogen activator

Dyes-based biomimetic affinity chromatography has been used to purify therapeutically useful proteins. In order to design novel biomimetic affinity ligands for purification of tissue-type plasminogen activator (t-PA), small molecular fragments were achieved to fit in S3/4 binding site of t-PA by structure-based ligand design method (InsightII/Ludi). Three biomimetic affinity ligands A, B, and C were then designed, synthesized, and proved to bind the target protein (t-PA), exceeding the binding capacity of the commercial p-amino benzamidine affinity matrix. The designed affinity matrix A showed high efficiency to purify sc-tpa from the crude samples with 18-fold of purification.

Urokinase separation from cell culture broth of a human kidney cell line

A single step ion-exchange chromatography on a sulfo-propyl (SP)- Sepharose column was performed to separate both the high molecular weight (HMW)- and low molecular weight (LMW)- forms of enzymatically active urokinase type plasminogen activator from human kidney (HT1080) cell culture media. The level of urokinase secreted by the cell line reached to about 145 Plough units/ml culture broth within 48 h of cultivation. The conditioned cell culture media was applied directly to the column without any prior concentration steps. Polyacrylamide gel electrophoresis of the column eluates in the presence of sodium dodecyl sulphate showed that the cell line secretes three forms of two-chain high molecular weight (HMW) urokinase of molecular weights (M(r)) 64,000, 60,900 and 55,000. In addition, two low molecular weight (LMW) forms of M(r) 22,000 and 20,000; proteolytic cleavage products of HMW, were also found. The HMW and LMW forms had intrinsic plasminogen dependent proteolytic activity as judged by zymographic analysis. The specific activity of the pooled peak fractions increased (approximately 93-fold) to values as high as 1481 Plough units/ mg protein. Both HMW as well as LMW forms were obtained in significantly high yields.

Expression of the non-glycosylated kringle domain of tissue type plasminogen activator in Pichia and its anti-endothelial cell activity

The two-kringle domain of tissue-type plasminogen activator (TK1-2) has been identified as a potent angiogenesis inhibitor by suppressing endothelial cell proliferation, in vivo angiogenesis, and in vivo tumor growth. Escherichia coli-derived, non-glycosylated TK1-2 more potently inhibits in vivo tumor growth, whereas Pichia expression system is more efficient for producing TK1-2 as a soluble form, albeit accompanying N-glycosylation. Therefore, in order to avoid immune reactivity and improve in vivo efficacy, we expressed the non-glycosylated form of TK1-2 in Pichia pastoris and evaluated its activity in vitro. When TK1-2 was mutated at either Asn(117) or Asn(184) by replacing with Gln, the mutated proteins produced the glycosylated form in Pichia, of which sugar moiety could be deleted by endoglycosidase H treatment. When both sites were replaced by Gln, the resulting mutant produced a non-glycosylated protein, NQ-TK1-2. Secreted NQ-TK1-2 was purified from the culture broth by sequential ion exchange chromatography using SP-sepharose, Q-spin, and UNO-S1 column. The purified NQ-TK1-2 migrated as a single protein band of approximately 20 kDa in SDS-PAGE and its mass spectrum showed one major peak of 19,950.71 Da, which is smaller than those of two glycosylated forms of wild type TK1-2. Functionally, the purified NQ-TK1-2 inhibited endothelial cell proliferation and migration stimulated by bFGF and VEGF, respectively. Therefore, the results suggest that non-glycosylated TK1-2 useful for the treatment of cancer can be efficiently produced in Pichia, with retaining its activity.

Glycosylation variant analysis of recombinant human tissue plasminogen activator produced in urea-cycle-enzyme-expressing Chinese hamster ovary (CHO) cell line

Tissue plasminogen activator (tPA) was produced in ornithine transcarbamoylase (OTC) cells by introducing the tPA gene into OTC cells. OTC cells were originally derived from Chinese hamster ovary (CHO) cells and express the first two enzymes of the urea cycle, carbamoyl phosphate synthetase I (CPS I) and OTC. To investigate glycosylation variants, tPA variants produced in serum-supplemented culture medium of OTC-tPA cells were separated by lysine-Sepharose 4B chromatography. Unlike in previous studies that used lysine-Sepharose chromatography, two peaks were identified to correspond to eluted glycosylation variants type I and II and type II and the percentages of the type I and type II variants were found to be 23% and 77%, respectively. The biological activities of the type I and II and type II variants were twofold that of the Third International tPA Standard (98/714) produced in the CHO cell line, and the activity of type II variant was 12.6% higher than that of the type I and II variants. These results demonstrate that tPA produced in urea-cycle-enzyme-producing OTC cells have a very high biological activity and the percentage of type II variant which is very valuable for the biopharmaceutical industry is higher than that of any report using CHO cells.

Preparation and characterization of scFv for affinity purification of reteplase

The work was to explore the feasibility of protein affinity purification using ligand isolated from phage library. Reteplase was used as the model protein and a humanized semi-synthetic single chain fragment variable phage library as the source of ligand. After four rounds of biopanning, reteplase-specific phage clones were greatly enriched. The scFv gene from the best phage clone was inserted to pET-29a and expressed in E. coli Rosseta. After purification by nickel-affinity and refolding, this scFv protein was proven to recognize reteplase specifically and sensitively in ELISA and dot-blotting. Its binding constant to reteplase was 1.84x10(-8) M, measured by surface plasmon resonance. After immobilized on Sepharose 4B, the scFv was used for the affinity purification of reteplase from milk. It was found that reteplase was highly purified from the starting material. In conclusion, it has been demonstrated that humanized scFv prepared with this approach could be used as a practical affinity ligand for efficient and cost-effective purification of reteplase, as well as other therapeutic proteins.

[Expression and activity of tissue-type plasminogen activator mutant reteplase with deletion of PAI-1 binding sites]

AIM

To construct a mutant of human tissue plasminogen activator (t-PA) with activity not inhibited by plasminogen activator inhibitor-1 (PAI-1).

METHODS

The finger, epidermal growth factor and kringle-1 domains were removed from native t-PA to obtain a t-PA mutant with the PAI-1 binding sites. A pUC18 plasmid containing the human t-PA full-length cDNA sequence was used as template, and the DNA sequences encoding 1-3 and 176-527 amino acids were amplified by PCR. The nucleotides AAG CAC AGG AGG (from 373 to 384) in the PAI-1 binding site changed to GCG GCC GCG GCG, so amino acid KHRR was replaced by AAAA correspondently.

RESULTS

Sequencing result showed that the above t-PA mutant DNA sequence was consistent with the expected. Then it was cloned in an E.coli expression vector and was highly expressed. The expressed protein occupied about 30% of total bacterial proteins as inclusion body. After denaturation and renaturation, the active t-PA mutant was obtained,activity of which was not inhibited by PAI-1.

CONCLUSION

This t-PA mutant may be developed to a new gene-engineering drug with higher therapeutic activity and lower dose requirement in the treatment of acute myocardial infarction and cerebrovascular thrombosis diseases.

Tissue-type plasminogen activator: a historical perspective and personal account

Over the past two decades tissue-type plasminogen activator (t-PA), the main physiological plasminogen activator, has been developed as a fibrin-specific thrombolytic agent for the treatment of various thromboembolic diseases. Milestones in this development include: first purification of human t-PA from uterine tissue, elucidation of the interactions regulating physiological fibrinolysis, thus providing a molecular basis for the concept of fibrin-specific plasminogen activation, first animal models of thrombosis and pilot studies in patients supporting the therapeutic potential of t-PA, cloning and expression of recombinant t-PA providing sufficient amounts for large scale clinical use, and demonstration of its therapeutic benefit in large multicenter clinical trials, mainly in patients with acute myocardial infarction (AMI), but also in patients with massive pulmonary embolism, ischemic stroke, deep vein thrombosis and peripheral arterial occlusion. Genetically modified variants of t-PA have been developed for bolus administration in patients with AMI.

Solid phase peptide synthesis on epoxy-bearing methacrylate monoliths

Monoliths based on a copolymer of glycidyl methacrylate (GMA) and ethylene dimethacrylate (EDMA) can be used directly as sorbents for affinity chromatography after solid phase peptide synthesis. The quality of the synthesized products, the amount of grown peptides on a support and the reproducibility of the process must be considered. A determination of the quantity of the introducing beta-Ala (and, consequently, the total amount of synthesized peptide) was carried out. Three peptides complementary to recombinant tissue plasminogen activator (t-PA) have been synthesized using Fmoc-chemistry on GMA-EDMA disks. The peptidyl ligands were analysed by amino acid analysis, ES-MS and HPLC methods. The affinity binding parameters were obtained from frontal elution data. The results were compared with those established for GMA-EDMA affinity sorbents formed by the immobilization of the same but separately synthesized and purified ligands. The immobilization on GMA-EDMA disks was realized using a one-step reaction between the amino groups of the synthetic ligand and the original epoxy groups of monolithic material. The affinity constants found for two kinds of sorbent did not vary significantly. Finally, the directly obtained affinity sorbents were tested for t-PA separation from a cellular supernatant.

Controlled shear affinity filtration (CSAF): a new technology for integration of cell separation and protein isolation from mammalian cell cultures

Controlled shear affinity filtration (CSAF) integrates animal cell separation and product isolation in a single unit operation through the use of a specifically designed rotating disk filter with incorporated membrane chromatography column. Because of the decoupling of shear force and pressure generation and the specific hydrodynamics of the system, shear rates can be easily optimized and precisely controlled to maximize filtration performance while viability of the shear sensitive animal cells is maintained. In this study, the general methodology is demonstrated using the integration of Chinese hamster ovary cell separation and isolation of recombinant tissue plasminogen activator (t-PA) as a model example. Direct capture of t-PA from cell culture broth was realized by using custom-made affinity membranes with lysine as a robust, small molecular weight affinity ligand. Small-scale t-PA adsorption experiments, as well as microfiltration experiments, were used to design the integrated CSAF process. A Chinese hamster ovary batch culture was processed with a lab-scale prototype, yielding 86% of the t-PA in the concentrated, particle-free eluate, whereas 95% of the bulk protein was removed. Because the viability of the cells is not significantly affected and high specific flux rates can be achieved, the CSAF technology should also be well suited for continuous perfusion with integrated product isolation. A truly continuous operation could be realized with two systems in tandem configuration.

Production of tissue plasminogen activator (t-PA) in Aspergillus niger

A protease-deficient strain of Aspergillus niger has been used as a host for the production of human tissue plasminogen activator (t-PA). In defined medium, up to 0.07 mg t-PA (g biomass)(-1) was produced in batch and fed-batch cultures and production was increased two- to threefold in two-phase batch cultures in which additional glucose was provided as a single pulse at the end of the first batch growth phase. Production was increased [up to 1.9 mg t-PA (g biomass)(-1)] by the addition of soy peptone to the defined medium. The rate of t-PA production in batch cultures supplemented with soy peptone (0.2 to 0.6 mg t-PA L(-1) h(-1)) was comparable to rates observed previously in high-producing mammalian or insect cell cultures. In glucose-limited chemostat culture supplemented with soy peptone, t-PA was produced at a rate of 0.7 mg t-PA L(-1) h(-1). Expression of t-PA in A. niger resulted in increased expression of genes (bipA, pdiA, and cypB) involved in the unfolded protein response (UPR). However, when cypB was overexpressed in a t-PA-producing strain, t-PA production was not increased. The t-PA produced in A. niger was cleaved into two chains of similar molecular weight to two-chain human melanoma t-PA. The two chains appeared to be stable for at least 16 h in culture supernatant of the host strain. However, in general, <1% of the t-PA produced in A. niger was active, and active t-PA disappeared from the culture supernatant during the stationary phase of batch cultures, suggesting that the two-chain t-PA may have been incorrectly processed or that initial proteolytic cleavage occurred within the proteolytic domain of the protein. Total t-PA (detected by enzyme-linked immunoassay) also eventually disappeared from culture supernatants, confirming significant extracellular proteolytic activity, even though the host strain was protease-deficient.

Expression of a novel chimeric protein containing the A chain of tissue-type plasminogen activator and the B chain of pro-urokinase in insect cells using the baculovirus system

A hybrid cDNA tu-pa, which contains Ser1-Thr263 of tissue-type plasminogen activator (t-PA) and Ser138-Leu411 of pro-urokinase (pro-UK) was constructed and expressed in the Sf9-AcNPV system. The expression level was approximate 2.5 mg/L. Tu-PA was purified via one-step affinity column conjugated with monoclonal antibody against the B chain of pro-UK, which showed a single band of approximate 60 kDa in SDS-PAGE. The specific activity of the chimeric protein on fibrin plate was 200,000 IU/mg protein. Tu-PA had a higher selectivity for fibrin than UK and pro-UK. Its activity can be promoted by CNBr degraded fibrin fragments as t-PA.

A refined kinetic analysis of plasminogen activation by recombinant bovine tissue-type plasminogen activator indicates two interconvertible activator forms

Bovine tissue-type plasminogen activator (tPA) was heterologously expressed in the methylotrophic yeast Pichia pastoris and characterized structurally and kinetically. The bovine single-chain tPA-mediated activation of bovine plasminogen was studied in the presence and absence of fibrinogen fragments. We have proposed a refined new method of kinetic analysis which allows examination of both stationary and prestationary phases of this process. The investigation revealed the presence of two interconvertible forms of the recombinant bovine tPA being in equilibrium at a 1 to 50 ratio. Only the minor form was able to bind and activate plasminogen. Saturation of the whole pool of tPA required high plasminogen concentration (Km >/= 5 microM) in order to reverse the equilibrium between the two forms. Fibrinogen fragments activated the single-chain tPA due to preferential binding and stabilization of the minor "active" form of the enzyme until all the molecules of tPA were converted. The same mechanism could be applied to human tPA as well. The Km values, obtained for recombinant bovine and human tPA in the presence of fibrinogen fragments, were found to be similar (Km = 0.1 microM) while kcat of human tPA was 5-10 times higher.

Interaction of human tissue plasminogen activator (t-PA) with pregnancy zone protein: a comparative study with t-PA-alpha2-macroglobulin interaction

Human pregnancy zone protein (PZP) is a major pregnancy-associated plasma protein strongly related to alpha2-macroglobulin (alpha2-M). Interactions of tissue plasminogen activator (t-PA) with PZP and alpha2-M were both investigated in vitro and the complexes were analyzed by polyacrylamide gel electrophoresis (PAGE). The results demonstrated that PZP-t-PA complex formation was evident within 1 h of incubation, whereas alpha2-M-t-PA complexes were formed after 18 h. Conclusions were supported by the following evidence: (i) PZP and alpha2-M complexes revealed changes of the mobility rate in non-denaturing PAGE, similar to those observed with alpha-Ms-chymotrypsin; (ii) both PZP and alpha2-M formed complexes of molecular size >360 kDa by SDS-PAGE, in accordance with the covalent binding of t-PA, which was previously reported for other proteinases; and (iii) PZP underwent a specific cleavage of the bait region with appearence of fragments of 85-90 kDa as judged by reducing SDS-PAGE. In contrast, the proteolytic attack on alpha2-M was found to occur more slowly, requiring several hours of incubation with t-PA for generation of an appreciable amount of fragments of 85-90 kDa. The appearance of free SH-groups of alpha-Ms was further investigated by titration with 5, 5'-dithiobis(2-nitrobenzoic acid). The maximal level of SH-groups raised was 3.9 mol/mol of PZP and 3.5 mol/mol of alpha2-M, indicating approximately one SH-group for each 180-kDa subunit. Finally, t-PA activity in PZP-t-PA complex was evaluated by measuring the hydrolysis of the chromogenic substrate Flavigen t-PA. Our results revealed that prolongation of the incubation period of this complex increased t-PA-mediated hydrolysis of Flavigen t-PA until a plateau was reached, approximately between 60 and 120 min. The present study suggests that PZP, by binding to t-PA, may contribute to the control of the activity of proteinases derived from fibrinolytic systems.

Investigation on fouling mechanisms for recombinant human growth hormone sterile filtration

During sterile filtration of recombinant human growth hormone solutions, severe membrane fouling was experienced compared to other protein preparations of significantly higher molecular weights and concentrations. This phenomenon was attributed to rhGH aggregation/adsorption occurring in the filter pore. To better understand this phenomenon, we examined several possible fouling mechanisms: (1) pore constriction, (2) adsorption due to nonspecific binding between protein and the membrane, (3) shear-induced adsorption, (4) hydrophobic interface-induced aggregation/adsorption. The protein solutions were sterily filtered using 0.22 mm filters, and their filtration fluxes were monitored. Filtration on the capillary and the noncapillary filters suggested that constraints by pore constriction and tortuosity played only a limited role. Filtration using filters with different degrees of protein binding tendency suggested that nonspecific adsorption was insignificant. The shear stress acting on the protein during filtration was small. RhGH which was intentionally sheared in a high-speed concentrically rotating device did not aggravate fouling tendency, suggesting that the shear-induced adsorption might not be the major fouling mechanism. The dynamic light scattering data showed a trace amount of rhGH aggregates always present in equilibrium with the hydrophobic (air-water and membrane-water) interface. These aggregates tended to be adsorbed to the membrane, and more aggregates were generated presumably due to the equilibrium between aggregates and protein monomers. This adsorption/aggregation process eventually fouled the membrane. When the hydrophobic interface was occupied by surfactant molecules, the equilibration kinetics ceased to generate aggregates, thereby minimizing membrane fouling. This study clarified the cause of such an unusual fouling phenomenon upon microfiltration.

Tissue plasminogen activator binding to the annexin II tail domain. Direct modulation by homocysteine

Tissue plasminogen activator binds to endothelial cells via the calcium-regulated phospholipid-binding protein annexin II, an interaction that is inhibited by the prothrombotic amino acid homocysteine. We sought to identify the tissue plasminogen activator binding domain of annexin II and to determine the mechanism of its modulation by homocysteine. Tissue plasminogen activator binding to immobilized annexin II was inhibited by intact fluid phase annexin II but not by its "core" fragment (residues 25-339). Two overlapping "tail" peptides specifically blocked 65-75% of binding. Localization of the tissue plasminogen activator binding domain was confirmed upon specific inhibition by the hexapeptide LCKLSL (residues 7-12). Expressed C9G annexin II protein failed to support tissue plasminogen activator binding, while binding to C133G, C262G, and C335G was equivalent to that of wild type annexin II. Upon exposure to homocysteine, annexin II underwent a 135 +/- 4-Da increase in mass localizing specifically to Cys9 and a 60-66% loss in tissue plasminogen activator-binding capacity (I50 = 11 microM). Upon treatment of cultured endothelial cells with [35S]homocysteine, the dithiothreitol-sensitive label was recovered by immunoprecipitation with anti-annexin II IgG. These data provide a potential mechanism for the prothrombotic effect of homocysteine by demonstrating direct blockade of the tissue plasminogen activator binding domain of annexin II.

Characterization of the acidic oligosaccharides assembled on the Pichia pastoris-expressed recombinant kringle 2 domain of human tissue-type plasminogen activator

The N-linked glycans assembled in Pichia pastoris on the recombinant kringle 2 domain of human tissue-type plasminogen activator (r-[K2tPA]) are composed of approx. 80% neutral and 20% charged species. After peptide:N4-(N-acetyl-beta-glucosaminyl)asparaginyl amidase-catalysed liberation of the oligosaccharides from the purified glycopeptide, the glycan mixture was resolved by HPLC on amino-silica-based resin. Oligosaccharide mapping of the resulting mixture by HPLC, gel filtration and time-of-flight matrix-assisted laser-desorption-ionization-with-delayed-extraction mass spectrometry (TOF-MALDI DE-MS) revealed that > 90% of the charged species consisted of a series of oligosaccharides possessing molecular masses that were consistent with a range of saccharides comprising phospho-Man10GlcNAc2-phospho-Man14GlcNAc2, with phospho-Man11GlcNAc2 representing the major species. The remaining material in the charged fraction contained identifiable phosphorylated glycans that were one or two mannose units shorter, and one to four mannose units longer, than those present in the above range of oligosaccharides. Treatment of the native charged glycan pool with alkaline phosphatase did not result in molecular-size alterations, showing that phosphomonoesters are not present. Mild acid hydrolysis of the glycans led to a decrease in the size of all charged glycans by one mannose residue, providing phospho-Man9GlcNAc2-phospho-Man13GlcNAc2. Following this procedure, treatment with alkaline phosphatase resulted in size decreases that were equivalent to the loss of one phosphate group from each glycan. This demonstrates that all charged glycans isolated contained phosphate in phosphodiester bonds to two mannose units. The present study shows that P. pastoris cells possess the capability of assembling phosphorylated glycans having the phosphate moiety present in phosphodiester linkages with two mannose units. These saccharides, like the neutral oligosaccharides, contain considerably smaller amounts of mannose than glycans present in other strains of yeast.

Renaturation and purification of recombinant tissue-type plasminogen activator expressed in E. coli

Under the control of trp promoter, human tissue-type plasminogen activator was expressed in E. coli in the form of inclusion body. The recombinant t-PA was recovered for renaturation from preparative native PAGE, gel by zinc acetate staining and electroelution. After renaturation in vitro, the recombinant t-PA was purified by benzamidine affinity chromatography and lysine affinity chromatography. The purified t-PA showed homogeneous on silver-stained SDS-PAGE gel, with a specific activity of 240,000 I.U./mg protein.

High-level secretion in Pichia pastoris and biochemical characterization of the recombinant kringle 2 domain of tissue-type plasminogen activator

The kringle 2 (K2) domain of tissue-type plasminogen activator (tPA) has been expressed in Pichia pastoris cell lines GSI 15 and KM71. This construct contained a hexahistidine sequence at the C-terminus of the kringle to aid in purification by immobilized metalion-affinity chromatography. The exact amino acid sequence of the isolated kringle was EAEAYV-[K2tPA]SR(H)6, where [K2tPA] represents amino acid sequence residues C1-C82 of the kringle domain (residues 180-261 of tPA). The clones of the yeast transformants provided large amounts of the recombinant (r)-[K2tPA]-containing polypeptide at levels that allowed ready purification of several hundred mg from shake flasks and near-gram levels from a high-biomass fermenter. Purification of the kringle domain directly from cell-conditioned media was accomplished in a single step by either immobilized Ni(+)-affinity chromatography or lysine-Sepharose affinity chromatography. N-linked glycans were present on approx. 30% of this yeast-expressed material, at N5 of the kringle (corresponds to N11 of the particular construct, N184 of full-length tPA). The expressed recombinant kringle recognized a conformation-specific monoclonal antibody generated against tPA that is directed to the K2 domain of the protein, interacted properly with various omega-amino acid ligands, and showed signature conformational properties when studied by differential scanning calorimetry and high-resolution 1H-NMR. The results demonstrate that the P. pastoris system can be employed to obtain large amounts of secreted and properly folded kringle domains.

Bovine erythrocyte haemolysates enhance plasminogen activation by tissue-type plasminogen activator

An active fraction that accelerates plasminogen activation by tissue-type plasminogen activator (t-PA) was purified from a haemolysate of bovine erythrocytes. When the haemolysate was mixed with t-PA, it produced a 2- to 3-fold increase in plasminogen activation as measured by an insoluble fibrinolytic assay system and a soluble amidolytic assay system with the chromogenic substrate S-2251. Zymographic analysis showed that, while the haemolysate increased t-PA activity, it did not alter the electrophoretic characteristics of the t-PA nor did it induce any fibrinolysis in the absence of t-PA or plasminogen. The haemolysate was devoid of plasmin and plasminogen activator activity but was most effective in accelerating plasminogen activation by t-PA in the presence of substrate. Based on the purification characteristics of the active fraction in the haemolysate, it appears to have a molecular weight of less than 10 kDa.

Inclusion bodies and purification of proteins in biologically active forms

Even though recombinant DNA technology has made possible the production of valuable therapeutic proteins, its accumulation in the host cell as inclusion body poses serious problems in the recovery of functionally active proteins. In the last twenty years, alternative techniques have been evolved to purify biologically active proteins from inclusion bodies. Most of these remain only as inventions and very few are commercially exploited. This review summarizes the developments in isolation, refolding and purification of proteins from inclusion bodies that could be used for vaccine and non-vaccine applications. The second section involves a discussion on inclusion bodies, how they are formed, and their physicochemical properties. In vivo protein folding in Escherichia coli and kinetics of in vitro protein folding are the subjects of the third and fourth sections respectively. The next section covers the recovery of bioactive protein from inclusion bodies: it includes isolation of inclusion body from host cell debris, purification in denatured state alternate refolding techniques, and final purification of active molecules. Since purity and safety are two important issues in therapeutic grade proteins, the following three sections are devoted to immunological and biological characterization of biomolecules, nature, and type of impurities normally encountered, and their detection. Lastly, two case studies are discussed to demonstrate the sequence of process steps involved.

Integrin-specific tissue-type plasminogen activator engineered by introduction of the Arg-Gly-Asp sequence

To tailor tissue-type plasminogen activator (tPA) to possess an affinity for the integrins, several mutants were constructed by introducing the Arg-Gly-Asp (RGD) sequene into the tPA molecule. These mutants were expressed in COS-1 cells and partially purified by lysine-Sepharose chromatography. The RGD-dependent binding of the mutants to platelet integrin, integrin alpha IIb beta 3, was evaluated by subtracting the nonspecific binding in the presence of 10 mM EDTA (or 1 mg/ml GRGDSP). The binding assay showed that two tPA mutants possess high affinity for the integrin in an RGD-dependent manner. One mutant is 148RGD-tPA with RGDS in place of DRDS (residues 148 to 151) in the loop region of the kringle 1 domain of tPA, and the other is 270RGD-tPA with RGDS in place of SQPQ (residues 270 to 273) in the linker region between the kringle 2 and protease domains. Using the chromogenic substrate Spectrozyme tPA, the 148RGD-tPA mutant was shown to possess amidolytic activity comparable with that of native tPA, while the 270RGD-tPA mutant exhibited several-fold lower activity. In addition, the 148RGD-tPA exhibited full tPA activity even when interacting with the integrin alpha IIb beta 3. These results suggest that the bifunctional 148RGD-tPA molecule might be useful as an improved thrombolytic agent specific for the platelet integrin, the integrin alpha IIb beta 3.

Capillary isoelectric focusing

Capillary isoelectric focusing is a useful analytical technique for characterization of protein mixtures and determination of protein isoelectric points. It is particularly useful in separation of protein glycoforms (Fig. 5), characterizing protein microheterogeneity (Fig. 6), and resolution of charge variants (Fig. 7). The capillary focusing process is analogous to conventional isoelectric focusing in gels, while the requirement for zone mobilization is unique to the capillary format with on-tube detection. A variety of mobilization methods have been described, and the selection of the mobilization method for a particular application depends on the capillary type, the instrument configuration, and the type of proteins to be analyzed. Capillary IEF is generally successful for proteins with a molecular weight up to about 150,000 that exhibit good solubility in aqueous buffers, but may be unsatisfactory for large or hydrophobic proteins. Because of precipitation and variation in mobilization efficiencies, use of internal standards is recommended in most applications. Capillary IEF can be compared to conventional gel IEF in terms of sample throughput and sensitivity. Conventional gels require approximately 4-6 hr to cast, run, and stain the gel, depending on whether silver or Coomassie staining is used. A typical gel contains 10 sample lanes, yielding a throughput of 25-35 min/sample. Capillary IEF separations (including focusing and mobilization) are typically 15-20 min. The mass sensitivity of conventional gel IEF is 36-47 ng for Coomassie staining and 0.5-1.2 ng for silver staining. In capillary IEF, sensitivity will depend on the volume of sample injected; assuming a capillary with a volume of 100 nl is completely filled with sample prior to focusing, the limit of detection will be approximately 1 microgram/ml or 0.1 ng injected. Thus capillary IEF compares favorably with conventional gel IEF in terms of detectivity and analysis time, and has the additional benefit of complete automation of the process including separation and data reduction.

Substrate specificity of tissue type plasminogen activator. Characterization of the fibrin independent specificity of t-PA for plasminogen

Tissue-type plasminogen activator (t-PA) is a remarkably specific protease: the only known substrate of this enzyme in vivo is a single peptide bond (Arg560-Val561) within the proenzyme plasminogen. Part of the substrate specificity of t-PA is due to a ternary interaction between fibrin, t-PA and plasminogen which reduces the Km of t-PA for plasminogen by a factor of 440. However, even in the absence of fibrin, t-PA continues to hydrolyze plasminogen more rapidly than does trypsin, a homologous serine protease. We have measured the extent of the specificity of t-PA for plasminogen by assaying t-PA and trypsin toward substrates modeled after the peptide sequence in plasminogen surrounding Arg560-Val561. Surprisingly, t-PA hydrolyzes these substrates with kcat/Km values which are 28,000-210,000-fold lower than those obtained using trypsin. Both the high activity toward plasminogen and the low activity toward peptides are also exhibited by the isolated protease domain. This suggests that the protease domain, in spite of its high homology to the nonspecific enzyme trypsin, is inherently specific for recognition of one or more structural features displayed by native plasminogen.

Plasminogen activators in tissue extract of aural cholesteatoma

Using a biochemical technique, the authors characterized and identified the plasminogen activator (PA) derived from tissue extracts of six aural cholesteatomas. The results of fibrin zymography indicated that the tissue extracts of two cholesteatomas demonstrated two lytic zones on fibrin-agarose plates. One of the lytic zones was at about 72 kd, while the other zone was at about 64 kd. Using various goat immunoglobulin G (IgG)-containing antibodies (anti-human uterine tissue type PA (t-PA), anti-human low-molecular-weight (LMW) urokinase, and nonspecific goat IgG) and plasminogen-free fibrin-agarose plates, we confirmed that the cholesteatoma tissue extracts contained 72 kd t-PA and 64 kd urokinase type PA (u-PA). Furthermore, we measured the t-PA and u-PA activities in the tissue extracts selectively by parabolic rate assay. In order to estimate the PA activity, we developed optimal conditions for this assay. The specific t-PA activity ranged from 0.03 to 0.43 mIU/micrograms-protein and the specific u-PA activity ranged from 0 to 0.35 mIU/microgram-protein. The highest percentage of u-PA with respect to the total PA activity was 44.9%. However, in four of the six cases, we failed to detect u-PA activity. In the present study, we thus clarified the presence of PAs in tissue extracts of aural cholesteatomas. Furthermore, we confirmed that measureable u-PA occurred in some tissue extracts. We anticipate that the u-PA in inflammatory tissues plays an important role in the degradation of the extracellular matrix via the formation of plasmin and collagenases.

The activation of type 1 and type 2 plasminogen by type I and type II tissue plasminogen activator

Tissue plasminogen activator (tPA) was fractionated using lysine-Sepharose affinity chromatography. Type I, type II, and a minor peak with high affinity for lysine (designated type D) tPA were recovered. In an indirect amidolytic assay involving native human Glu-plasminogen and fibrin, type II tPA showed a 2-fold higher activity than type I. To explore the combinatorial effect of the variable glycosylation status of both tPA and plasminogen, kinetic constants for fibrin-dependent plasminogen activation were determined for combinations of type I, II, and D tPA with type 1 and 2 plasminogen. Within a 4-fold range, the fastest rate was achieved from the combination of type D (type II + D) tPA and type 2 plasminogen. N-Glycosylation of plasminogen increased the Km value for activation by all tPA variants; N-glycosylation of type I tPA at Asn184 decreased the kcat (turnover) values for the fibrin-dependent activation of plasminogen over type II tPA, while type D tPA showed the highest turnover rate. In the presence of fibrinogen fragments, N-glycosylation of plasminogen at site 289 modulates the kinetics of association of enzyme and substrate, while N-glycosylation at site 184 on tPA modulates the turnover rate of the enzyme.

Selectivity of the plasminogen activator inhibitor (PAI-1) for the iso enzyme of guanidinobenzoatase on the surface of colonic carcinoma cells

The interaction of plasminogen activator-inhibitor (PAI-1) with a cell surface protease, guanidinobenzoatase (GB), has been studied in free solution and on the surface of colonic epithelial cells. It has been demonstrated that PAI-1 recognises and inhibits the iso enzymic form of GB associated with colonic carcinoma cells but fails to bind to the iso enzymic form of GB associated with normal donor colonic epithelial cells. This interaction is mediated by a lysyl binding site on the GB: complex formation prevents GB binding to fibrin fibrils which also involves lysyl binding sites.

Characterization of the murine plasma fibrinolytic system

The main components of the murine plasma fibrinolytic system, including fibrinogen, plasminogen, alpha 2-antiplasmin, tissue-type plasminogen activator and plasminogen activator inhibitor-1, were purified to homogeneity and their interactions were quantitated and compared with those of the human counterparts. Initial activation rates of murine and human plasminogen by autologous tissue-type plasminogen activator were comparable (catalytic efficiencies, k2/Km, of 0.4 and 0.6 mM-1 s-1, respectively), but murine plasminogen appeared to be resistant to activation by human tissue-type plasminogen activator (k2/Km = 0.01 mM-1 s-1). Plasminogen activation by tissue-type plasminogen activator was stimulated 100- and 160-fold in autologous murine and human systems, respectively, with saturating concentrations of 0.45 and 0.32 microM, respectively, of CNBr-digested fibrinogen. Nearly quantitative binding (85-90%) of tissue-type plasminogen activator to fibrin was observed both in autologous and heterologous systems. Murine and human plasmin were very rapidly inhibited by autologous and heterologous alpha 2-antiplasmin (second-order inhibition rate constants, k1,app, of 2.1-2.3 x 10(7) M-1 s-1) and murine and human tissue-type plasminogen activator were very rapidly inhibited by autologous or heterologous plasminogen activator inhibitor-1 (k1,app of 1.8-4.9 x 10(7) M-1 s-1). Two-chain murine tissue-type plasminogen activator (added at a concentration of 1 microgram/ml) was inhibited in normal or plasminogen activator inhibitor-1-deficient murine plasma with half-lives of 6.5 min and 4.2 min, respectively, as compared to 80 min for human tissue-type plasminogen activator, suggesting that murine plasma contains proteinase inhibitors other than plasminogen activator inhibitor-1 which efficiently inhibit autologous tissue-type plasminogen activator. Clot lysis experiments in autologous plasma revealed that the murine plasma fibrinolytic system is more resistant to activation than the human system (20-30% clot lysis in 2 h with 100 nM tissue-type plasminogen activator in the murine system, as compared to 50% clot lysis in 2 h with 3.5 nM tissue-type plasminogen activator in the human system). Several mechanisms appear to be involved in this relative resistance observed in the murine system, including resistance of murine plasminogen to quantitative activation and short plasma half-life of murine tissue-type plasminogen activator. Thus, although these quantitative interactions between purified components of the murine fibrinolytic system appear to be comparable to those between the human counterparts, murine plasma clots are > 30-fold more resistant to lysis with autologous tissue-type plasminogen activator than human plasma clots.

Efficiency and economics of bioseparation: some case studies

An efficiency and economic analysis of the bioseparation of different bioproducts is presented by utilizing different case studies. Economic data on bioseparation processes is rarely presented in the open literature. The different examples presented together attempt to provide an overall framework for the cost structure involved in the different bioseparation processes. The models presented (Datar, 1986; Porter and Ladisch, 1992) provide novel insights into the cost structure of the different aspects of bioseparation processes. More studies are required that provide the insights necessary to assist in improving the efficiency and economics of bioseparation processes.

Purified protein products of rDNA technology expressed in animal cell culture

During the past 7 years, 14 versions of 7 rDNA proteins have been licensed which are derived from animal cell culture expression systems. These medically useful products have included hormones, coagulation factors, enzymes and a vaccine. Aspects of the molecular complexity, manufacture, control and utilization of these products are discussed. In contrast to previous generations of biological production technology, the technology for production of rDNA-derived proteins in animal cells appears to be safe.

Prostacyclin is required for t-PA release after venous occlusion

The role of vascular cyclooxygenase pathway on tissue-type plasminogen activator (t-PA) release after venous occlusion was studied in anesthetized rats. After the inferior vena cava was clamped for 30 min, fibrinolytic activity increased from 143.7 +/- 14.5 to 209.5 +/- 10.3 mm2 (mean +/- SE, P < 0.002). This increase was prevented by aspirin at high (100 mg/kg i.v.) but not at low doses (1 mg/kg i.v.). Dazoxiben (10 mg/kg i.v.), an inhibitor of thromboxane synthase, was ineffective on the fibrinolytic response. Both the basal levels of 6-ketoprostaglandin F1 alpha and its increase after venous occlusion were suppressed by 100 mg/kg aspirin administration (from 0.64 +/- 0.2 to 0.05 +/- 0.002 ng/ml before occlusion, P < 0.001; and from 1.08 +/- 0.2 to 0.06 +/- 0.002 ng/kg after occlusion, P < 0.001), whereas they were both unaffected by aspirin at low doses (from 0.53 +/- 0.06 before to 1.20 +/- 0.08 ng/ml after stasis). Moreover, iloprost, a stable analogue of prostacyclin, reversed the aspirin inhibitory effects on fibrinolytic activity by restoring t-PA vascular release after venous stasis. Our results provide experimental evidence that an intact cyclooxygenase pathway in vascular wall is required for the fibrinolytic activity increase after venous occlusion in rats.

Sodium dodecyl sulfate-induced dissociation of complexes between human tissue plasminogen activator and its specific inhibitor

The stability of complexes between serine proteinases and their inhibitors after sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis has been claimed to indicate covalent bond formation. In this work we have investigated the effects of SDS on the stability of complexes between single-chain or two-chain tissue plasminogen activator (t-PA) and its inhibitor (PAI-1). Complexes formed by incubation of t-PA with PAI-1 for 15 min at 22 degrees C were further incubated with various amounts of SDS before being subjected to SDS-polyacrylamide gel electrophoresis. The molecular species in the gels were identified both by zymography or by autoradiography after immunoblotting with antibodies directed against either t-PA or PAI-1. It was demonstrated that the interaction of SDS with t-PA.PAI-1 complexes before electrophoresis resulted in a transition from the complexed state to the free forms of t-PA and PAI-1 in a time- and dose-dependent manner. The first-order dissociation rate constant in the presence of 35 mM SDS at 22 degrees C had a koff value of 1.4 x 10(-2) min-1, which corresponds to a half-life of 49.5 min. The t-PA released from the complexes was fibrinolytically active, whereas the released PAI-1 inhibited activator-dependent fibrinolysis. In a similar fashion, the well characterized nonacylated pair alpha 1-proteinase inhibitor-elastase was dissociated by SDS treatment, confirming the validity of our experimental approach to demonstrate the reversibility of t-PA.PAI-1 complexes. These results demonstrate that SDS-polyacrylamide gel electrophoresis traps the molecular species in the state in which the proteins existed prior to the analysis, and they suggest that under the conditions used, the interaction of t-PA with PAI-1 results in the formation of nonacylated reversible complexes. This phenomenon may be relevant to the pathophysiology of fibrinolysis and to the general mechanism of serine proteinase-inhibitor complex formation.

Regulation of plasminogen activation in rat cell lines

The regulation of plasminogen activators (PA) and their inhibitors (PAI) in the rat cell lines: HTC and L2 was studied. HTC plasminogen activator inhibitor type 1 (PAI-1) production was stimulated by dexamethasone, serum factors and insulin; that of tissue-type plasminogen activator (tPA) by cAMP raising agents. Retinoic acid, butyrate, phorbol ester and endotoxin did not affect net PA/PAI activity elaborated by HTC. L2 cells produced tPA, which production was stimulated by retinoic acid, phorbol myristate acetate, butyrate and cAMP; serum factors blunted their response, whereas in the synthetic serum substituting medium Ultraculture and with cocktail Ultroser the action of tPA stimulators was enhanced.

Correlation of in vitro and in vivo decreased fibrinolytic activity caused by dexamethasone

The effect of dexamethasone administration to rats was studied on blood fibrinolytic activity. PAI-1 levels were dose-dependently enhanced by dexamethasone after a pretreatment period of 5 days while simultaneously a decreased tPA level was observed. These ex vivo measured effects were confirmed in vivo with a specially developed fibrinolysis model. It is concluded that the in vivo measured inhibition of the fibrinolytic activity caused by dexamethasone correlates well with ex vivo measured activities.

Endocytosis and lysosomal delivery of tissue plasminogen activator-inhibitor 1 complexes in Hep G2 cells

Receptor-mediated endocytosis of tissue-type plasminogen activator (t-PA)-plasminogen activator inhibitor type 1 (PAI-1) complexes results in their clearance by Hep G2 cells. After complexes are internalized, the t-PA component is degraded. However, neither the locus of intracellular catabolism nor the fate of PAI-1 has been elucidated. To characterize these aspects of t-PA-PAI-1 catabolism, the subcellular distribution of a prebound cohort of ligand molecules was delineated after internalization at 37 degrees C. 125I-t-PA.PAI-1 and t-PA.125I-PAI-1 were compared in separate experiments. After ligand uptake, intracellular vesicles were separated on density gradients. Internalized 125I-t-PA.PAI-1 concentrated initially in endosomes. After 20 minutes of uptake, the complex began to appear in lysosomes. Subsequently, low molecular weight labeled ligand fragments were detected in culture media. A panel of lysosomotropic agents, including primaquine, chloroquine, ammonium chloride, and a combination of leupeptin and pepstatin A, inhibited degradation. When t-PA.125I-PAI-1 rather than 125I-t-PA.PAI-1 was internalized, strikingly different results were observed. Although the kinetics of internalization and the intracellular itinerary were indistinguishable for the differently labeled complexes, the 125I-PAI-1 component of t-PA.125I-PAI-1 resisted rapid degradation. After a rapid loss of t-PA, the 125I-PAI-1 moiety persisted in lysosomes for up to 180 minutes. Thus, internalized t-PA.PAI-1 is targeted to lysosomes in which PAI-1 is relatively more stable than t-PA.

One-step purification of tissue-type plasminogen activator using affinity chromatography with a special monoclonal antibody under mild conditions

We have previously isolated a monoclonal antibody, designated as 1-3-1, specific for tissue-type plasminogen activator (t-PA). We have shown that t-PA dissociates from 1-3-1 in the presence of the lysine analogue 6-aminohexanoic acid (6-AHA). Here we describe a method for the one-step immunoaffinity purification of t-PA from conditioned melanoma cell medium, using 1-3-1 immobilised on Sepharose under mild elution conditions, favourable for t-PA. The yield of t-PA (antigen or total protein) from a 1-3-1-Sepharose column, when eluted using a buffer supplemented with 0.2 M 6-AHA at neutral pH, was as effective as other buffers that involve a strong pH-change, i.e., pH 2-3. However, the enzymatic activity of the t-PA purified with 6-AHA was 25 to 30% higher, as compared with t-PA eluted using a pH change. This resulted in a markedly higher specific activity of t-PA purified with 0.2 M 6-AHA, as compared with t-PA purified using a strong pH-change. The purity of t-PA, purified using the present method, was very high, as determined by gel electrophoresis. An additional advantage of the present procedure is that the mild elution conditions prolong the column life.

Cell-free synthesis of enzymically active tissue-type plasminogen activator. Protein folding determines the extent of N-linked glycosylation

Tissue-type plasminogen activator (t-PA) is synthesized in mammalian cells as a mixture of two forms that differ in their extent of N-linked glycosylation. We have investigated the mechanism underlying this variation in glycosylation, using a cell-free system that consists of a rabbit reticulocyte lysate optimized for the formation of disulphide bonds and supplemented with dog pancreas microsomal membranes. Molecules of human t-PA synthesized in vitro are enzymically active and responsive to natural activators and inhibitors, and are glycosylated in a pattern identical with that of the protein produced in vivo. This demonstrates that t-PA synthesized in vitro folds into the same conformation as the protein synthesized in vivo. We show that the extent of glycosylation of individual t-PA molecules is dependent on the state of folding of the polypeptide chain, since the probability of addition of an oligosaccharide side chain at Asn-184 is decreased under conditions that promote the formation of enzymically active molecules. This variation in glycosylation is independent of the rate of protein synthesis.

Low density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor is an hepatic receptor for tissue-type plasminogen activator

Tissue-type plasminogen activator (t-PA), a serine protease that catalyzes the initial and rate-limiting step in the fibrinolytic cascade, is cleared rapidly in vivo by the liver. Using chemical crosslinking, we have recently identified a plasminogen-activator inhibitor type 1 (PAI-1)-independent t-PA clearance receptor on rat hepatoma MH1C1 cells with a relative molecular mass of approximately 500 kDa. Another recently identified membrane receptor, low density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor (LRP/alpha 2MR), was also detected on MH1C1 hepatoma cells by using immunoprecipitation with anti-LRP/alpha 2MR antibody. When analyzed by SDS/PAGE, we found the t-PA receptor identified on MH1C1 cells comigrated with the large subunit of LRP/alpha 2MR. The t-PA receptor was immunoprecipitated by an anti-LRP/alpha 2MR antibody after chemical crosslinking of specifically bound 125I-labeled t-PA to its receptor. Through chemical crosslinking studies, we found that t-PA and methylamine-activated alpha 2-macroglobulin could bind to LRP/alpha 2MR simultaneously without competing with one another for binding, suggesting that the two ligands bound to two independent sites on the LRP/alpha 2MR molecule. Furthermore, a 39-kDa protein, which modulates ligand binding to LRP/alpha 2MR, was also found to inhibit t-PA binding to its receptor. These data thus show that the t-PA clearance receptor identified on MH1C1 hepatoma cells is LRP/alpha 2MR.

Invasion of brain tissue by primary glioma: evidence for the involvement of urokinase-type plasminogen activator as an activator of type IV collagenase

The immunocharacterization of a metalloproteinase isolated from rat glioma cell conditioned medium is described and confirms that the enzyme is identical to type IV collagenase. Free, active plasminogen activator (PA) and PA-PAI complexes were identified as being secreted by the same cells. Using affinity-purified metalloproteinase we demonstrate that the enzyme can be partially activated by u-PA but not by plasmin in vitro. On the basis of these findings and previous published work we propose a scheme for the proteolytic degradation of normal brain tissue during tumour invasion.