Purification and Characterization of Microbial Heparinase for Production of Low Molecular Weight Heparins (LMWHs) Acting as Antithrombotic Agents

Abstract

Guide- Dr. P. Kalpana Murthy, Ph D Thesis Submitted to Chhatrapati Shahu Ji Maharaj University Kanpur in 2010.

Soil samples from basin of Gomti river (Lucknow, 26050’27”N, 80056’48”E), planes of northern India (Amritsar, 31037’59”N, 74051’56”E; Midnapore, 22025’0”N, 87019’0”E), and from Himalyan mountain region (Kedarnath, 30044’1”N, 7904’0”E) were collected and screened for potential heparinise producing microorganisms by selective enrichment culturing in heparinise selective medium having heparin (1% w/v) as the sole carbon and nitrogen source. Selection of the microorganisms on the basis of growth in the heparinise selective medium led to the isolation of twenty four bacterial and fungal cultures. Secondary screening on hepariase plate assay led to the separation of three fungal and two bacterial strains, which showed marked utilization oh heparin, as observed by clearance zone around the microbial growth against protamine sulphate precipitation. The axenic cultures of the five screened microorganisms were subjected to shake flask fermentation, for the selection of the microorganism on the basis of heparinise production in heparnaseselective medium. Out of five microorganisms screened for the heparinise production, two isolates, one fungal (H4) and one bacterial (B1), showed high heparinise production of 394.1 U/L and 342 U/L, respectively. The identification and characterization of the heparinise producing microbial isolates was done morphologically, physiologically as well as by genetic homology studies. The fungal strain (H4) was identified as Aspergillus flavus by 26S rRNA gene homology and thebacterial strain (B1) was identified as Acinetobacter calcoaceticus by 16S rRNA gene homology studies. Heparinase enzyme production in submerged batch fermentation conditions at shake flask level was studied for both the isolates in the defined heparinise minimal medium (HMM) and complex media (basal production medium, BPM and BPM devoid of heparin). Heprain dependent induction of heparinise production in the early log phase of growth (heparin utilization rate, 0.16 g/g cell/hour) was observed for A. Flavus in BPM and enzyme production remained repressed in BPM devoid of heparin. On the contrary, hearin independent induction of heparinise at 35.2 U/L was observed for A. Calcoaceticus in BPM devoid of heparin. In HMM, where heparin was the sole nitrogen source, significantly lower growth rates (max, 0.11 hour-1 and 0.05 hour-1) was observed as compared to that of BPM (max, 0.13 hour-1 and 0.07 hour-1), for A. Flavus and A. Calcoaceticus, respectively. A maximum cellular protein concentration of 394.2 g/mL was found for A. Calcoaceticus at 24 hours of growth in BPM as compared to 143.1 g/mL after 32 hours of growth in HMM. Similarly, maximum cellular protein concentration of 484.9 g/mL was found for A. Flavus in BPM as compared to the protein concentration of 175.4 g/mL in HMM after 58 hours of growth. Despite the slow growth of the A. Flavus and A. Calcoaceticus in the HMM, heparinise specific activity was high (242.33 U/mg protein and 323.35 U/mg protein, respectively) in comparison to heparinise specific activity (98.47 U/mg protein and 152.03 U/mg protein, respectivelt) in BPM. Medium optimization was done in basal production medium where heparin (0.2 g/L) was added as an inducer in the presence of dextrose (8 g/L), soy peptone (12.5 g/L) and ammonium sulphate (2 g/L) as alternative carbon and nitrogen sources. In classical one factor at a time, shake flask level, experiments for A. Flavus, replacement of glucose to mannitol, soy peptone to chtin and ammonium sulphate to ammonium nitrate in the basal production medium enhanced heparinise production by 94.83% from 387 U/L to 754U/L. Classical medium optimization experiments for A. Calcoaceticus resulted in 33.33% increase in heparinise production, from 72U/L to 54 U/L, where glucose, casein acid hydrolysate and ammonium bicarbonate were alternative carbon and nitrogen sources. Statistical optimization of medium components by two step process involving Plackett-Burman design (PBD) and central composite design along with response surface methodology were employed for economic and effective production of heparinise at shake flask level. Interaction of variables at two levels in PBD revealed that ammonium nitrate, chitin and heparin were significantly affecting (p 0.001-0.03) the heparinise yield of A. Flavus and ammonium bicarbonate, glucose, casein acid hydrolysate and heparin were most significantly affecting (p 0.0001-0.0049) the heparinise yield of A. Calcoaceticus. The heparinise production from A. Flavus, in classically optimized medium was 754 U/L and after statistical optimization increased by 1.37 fold to 1784 U/L. Optimization experiments as a whole (classical and statistical) increased the heparinise yield by A. Cadcoaceticus from 54 U/L to 186.8 U/L, a 3.5 fold increase. Heparinase and biomass production by A. Flavus and A. Calcoaceticus was compared in different flasks presenting various shapes while keeping the overall volume of each flask as the same (500 mL). Under submerged fermentation, at 180 rpm orbital shaker conditions (volumetric oxygen transfer coefficient, kLa, 6.5 hour-1) in the optimized production medium, the bacterium showed slow growth (generation time 91.29 minutes) and 186.8 U/L heparinise production. Enhanced agitation and aeration (kLa, 31.5 hour-1) in flask with four baffles at the base resulted in bacterial growth at 48.16 minutes generation time and corresponding 297.81 U/L heparinise yield. On the contrary, variation in agitation and aeration brought about by baffles was detrimental for heparinise and biomass production by A. Flavus. Agitation and aeration requirement of both the isolates was investigated sequentially in bubble column bioreactor and stirred tank bioreactor. It was observed that for A. Flavus biomass and intracullar heparinise production was associated with the pellet morphology and the culture reqired lower agitation and aeration rates. Agitation changed the compact pellet morphology to coagulative type to broken and excessive loose mycelia at higher agitation. The optimization of fermentation conditions (100 rpm agitation and 0.8 vvm aeration) led to 1954 U/L heparinise production in 1 L stirred tank Gallenkamp bioreactor, a mere 9.5% increase from the shake flask fermentation (1784 U/L). The kinetic parameters of the heparinise production process, in Gallenkamp bioreactor at 100 rpm agitation and 0.8 vvm aeration by A. Flavus and under optimized fermentation conditions in the Bioflo 110 bioreactor by A. Calcoaceticus were obtainedby applying logistic growth kinetics model, the Luedeking-Piret product formation model and the substrate utilization model. Positive value for  and  Luedeking-Piret model parameters (, 349.76 U/g cell/hour) by A.flavus indicated that the enzyme formation was growth associated. The medium optimization as well as modelling of the growth and enzyme production parameters led to the development of a suitable process for large scale production of heparinises reqired for its purification. Since both the organisms were producing intracellular heparinise enzyme, the first step of purification was preparation of cell extract. A three cycle freeze thaw procedure (liquid nitrogen and 370C water bath) was employed for partial cell lysis of the outer lipid layer disruption of Gram negative A. Calcoaceticus. The preferential release of periplasmic heparinise by freeze thaw resulted in high heparinise specific activity in the cell extract and exclusion of the precipitation of nucleic acid by protamine sulphate. Conversely, the mycelia suspension was disrupted by sonication and the nucleic acids were precipitated by protamine sulphate at a concentration of 1 mg/mg protein. Further purification of heparinises was achieved by anion and cation exchange liquid chromatography based on the typical physical characteristic of all the known heparinises from various microbial isolates, which show isoelectric point in the alkaline range of 8.5- 9.9. Finally, gel permeation chromatography led to the separation of apparently pure heparinise enzymes, as indicated by SDS-PAGE, of 120,000 and 24, 000 Dalton from A. Calcoaceticus and A. Flavus, respectively. The relative heparinise activity of A. Calcoaceticus was maximum at 350C in the presence of 250 mM NaC1 at pH 7.5 and was inhibited in the presence of Ba2+, Hg2+, Cd2+. Contrary to the heparinise from A. Calcoaceticus, the activity of heparinise from A. Flavus was higest in the presence of Co2+ and Mn2+, but inhibited by Pb2+ , Cd2+, Ba2+, and Zn2+ ions. The optimal heparinise activity for A. Flavus was found at pH 7.0 and the heparinise was highly active within the temperature range of 25 to 350C, and NaC1 concentration of 300 mM. Cysteine modifier (iodoacetic acid, IAA) and histidine modifier (diethylene pyrocarbonate, DEPC) inhibited the activity of both the enzymes, which suggest that cysteine and histidine, like that of heparinise I & II from Pedobacter heparinus, may be present at the active site of these heparinises. In addition, both thiol reducing reagents (dithiothreitol, -mercaptoethanol) enhanced the enzyme activity suggesting that the disulfide linkage is not essential for the enzyme activity, which is further supported by the decrease in enzyme activity in the presence of IAA, a carboxymethylating agent of the sulfohydryl group of cysteine. Both the heparinises were able to degrade heparin of porcine and bovine origine, but to a limited extent heparin sulphate, indicating their catalytic similarity with heparinise I of Pedobactor heparinus. The Km for porcine heparin as substrate for the heparinises from A. Calcoaceticus and A. Flavus were calculated as 2.610-5 M and 2.210-5 M, respectively. Investigations were made towards employing various means to utilize microbial physiology for heparin depolymerisation. The mass transfer limitations imposed by the microbial cell wall, resulted in less effective contact between substrate and active enzymes, and hence higher heparin depolymerisation retes were observed with the soluble heparinise. The process of batch depolymerisation was monitored for various parameters like uronic acid absorbance at 232 nm, heparin metachromasia, viscosity measurement, gradient PAGE analysis, fractionation by gel permeation and activated partial thromboplastin time (aPTT). Oligosaccharides formed at different time intervals, as fractionated by Sephadex G-50 column, revealed that in initial 16-24 hours of depolymerisation, 30-40% of heparin, of porcine origin with average molecular weight of 12000 Da, was cleaved to produce middle range (6000-3000 Da) heparin oligosaccharides. As the deploymerization of heparine proceeded to 48 hours, 85-95% of the heparin was cleaved to produce disaccharide as the main product. It was also found that at 30 % enzymatic depolymerisation of heparin; with the formation of oligosaccharides of 6000 Da average molecular weight, the heparin anticoagulant (aPTT) activity decreased by 58-80%. The deep concave curved plot of initial viscosity versus percent of final absorbance at 232 nm of heparin depolymerisation indicated that the heparinises from A. flavus and A. Calcoaceticus cleave the heparin in a random, endolytic fashion. Gradient PAGE of heparin deploymerized with membrane entrapprd heparinises from A. Flavus and A. Calcoaceticus, along with heparinise I from P. Heparinus supported the random endolytic action pattern of heparinises. As the depolymerisation of heparin continued, the higher molecular weight oligosaccharides were depolymerised with simultaneous appearance of oligosaccharides of degree polymerization <10. The membrane entrapped purified heparinise from A. Calcoaceticus; A. Flavus and P. Heparinus were used for generation of low molecular weight heparins (active oligosaccharides) by incomplete depolymerisation (< 30%) of heparin in a bubble column bioreactor for 24 hours. Heparin deploymerization was regulated by measured increase in light absorption at 232 nm, an increase which was related to the average molecular weight of heparin degradation products in the reaction mixture. The different fraction s of oligosaccharides were separated by size exclusion chromatography (Sephadex G-50) and evaluated for antithrombotic activity in terms of inhibition of blood coagulation factor Xa and IIa (anti-Xa and anti-IIa). The fraction having molecular weight in the range 6000-5000 Da were found to have anti-Xa:anti-IIa ratio >1.5, which consolidated the affect of oligosaccharide chain length on the antithrombotic activity of LMWHs.