Irreversible Immobilization of Diisopropylfluorophosphatase in

Géraldine F. Drevon, Claus Urbanke, and Alan J. Russell ... Géraldine F. Drevon, Judith Hartleib, Eileen Scharff, Heinz Rüterjans, and Alan J. Russ...
7 downloads 0 Views 155KB Size
Download PDF
Biomacromolecules 2000, 1, 571-576

571

Irreversible Immobilization of Diisopropylfluorophosphatase in Polyurethane Polymers Ge´ raldine F. Drevon and Alan J. Russell* Department of Chemical and Petroleum Engineering & Center for Biotechnology and Bioengineering, University of Pittsburgh, 1249 Benedum Hall, Pittsburgh, Pennsylvania 15261 Received February 15, 2000; Revised Manuscript Received June 15, 2000

The synthesis of polyurethane polymers in the presence of diisopropylfluorophosphatase (DFPase) has enabled the irreversible attachment of the enzyme to the polymeric matrix. The resulting bioplastic hydrolyzes diisopropylfluorophosphate (DFP) in buffered media up to 67% of the rate for the same amount of soluble enzyme. Above a DFPase concentration of approximately 0.1 mg/gfoam, the rate of the reaction catalyzed by the enzyme-containing polymer was controlled by internal mass transfer. Increasing foam hydrophilicity, via the use of nonionic surfactants during polymerization, significantly affected the structural properties of matrix, thereby enhancing the intrinsic and apparent efficiency of modified DFPase. The resulting reduction in internal mass transfer limitations was explained morphologically with electron microscopy. Introduction The existence of a stockpile of organophosphorus compounds across the globe, along with the threat of chemical warfare, leads to the necessity to develop strategies for the protection of individuals. Biocatalytic decontamination of nerve agents is now a leading tool in this growing list of potential responses to an incident.1 Biocatalytic decontamination is particularly attractive since it constitutes an environmentally benign approach with high specificity and efficiency under ambient conditions. Our research group has focused on methods to overcome perceived shortcomings of the enzymatic approach for the past decade. Disadvantages such as short lifetime, limited solubility in organic solvents, and expensive separation from reactants and products can be overcome via immobilization of agentases (enzyme which degrade nerve agents).2-4 Incorporation into a polymer network through multipoint attachment is a rapid and effective general strategy for enhancing the stability of enzymes, while retaining activity.5-9 This strategy involves the production of bioplastics in a single step, employing monomers capable of chemical reaction with specific functionalities on the enzyme surface. The covalent binding provides protein retention in the matrix so that the advantages of immobilization can be maximized. In a related study, we have recently measured the degree to which isocyanates react with proteins. Initially, work in our own laboratory focused on the incorporation of subtilisin into polyacrylates using nonconventional media.4,10,11 The method was then extended to organophosphorus hydrolase (OPH) immobilization into polyurethane foam.12-14 OPH is an effective catalyst for the degradation of organophosphorus nerve agents, which are often characterized by a low solubility in aqueous media. The polyurethane polymer represented an attractive support material, since its relative hydrophobicity facilitates the * Corresponding author. 300 Technology Drive, Center for Biotechnology & Bioengineering, University of Pittsburgh, Pittsburgh, PA 15219. Telephone: (412)-383-9740. Fax: (412)-383-9710. E-mail: [email protected].

interaction between biocatalyst and hydrophobic substrates. The biopolyurethanes also exhibit attractive chemical and mechanical properties such as solvent resilience and structural flexibility. OPH was immobilized during polyurethane foam synthesis, which involves the reaction of water with the multiple isocyanate functionalities of a foamable polyurethane prepolymers, Hypol.15 Hypol prepolymers are either aliphatic or aromatic polyisocyanate-terminated polyethers. The reaction between isocyanate groups and water molecules results in the formation of unstable carbamic acid, which decomposes into primary amine and carbon dioxide. This degradation initiates the foaming process with the carbon dioxide bubbling through the reacting system, providing the porosity of the polyurethane polymer. The primary amines produced during the first step of polymerization compete with water to further react with isocyanates, leading to the formation of a cross-linked polyurethane matrix. The protein participates to the foam synthesis via its lysine residues, enabling the production of a theoretical enzyme-polymer network. The addition of surfactants during polymerization changes the surface tension between reacting polymer and released carbon dioxide, influencing the carbon dioxide evolution and hence affecting the physical properties of the foam matrix, such as density, surface area, porosity, and rate of wetting. In the study reported herein, we describe the immobilization of diisopropylfluorophosphatase (DFPase, EC 3.8.2.1) into polyurethane matrices. DFPase catalyzes the hydrolysis of toxic organophosphorus compounds such as soman.16,17 Since alterations in foam morphology could influence the enzymatic activity retention, we performed the immobilization process in the presence and absence of Pluronic surfactants L62, P65, and F68 (from BASF). These nonionic surfactants are copolymers of ethylene oxide and propylene oxide with hydroxyl termini. The activity of soluble DFPase is dependent on ionic strength, and we therefore investigated the activity of the bioplastics in the absence and presence of

10.1021/bm0000034 CCC: $19.00 © 2000 American Chemical Society Published on Web 09/08/2000

572

Biomacromolecules, Vol. 1, No. 4, 2000

salt (NaCl). In each case, the degree to which the enzyme was irreversibly attached to the support was determined. The influence of mass transfer on the activity of enzymepolymers, as well as the effect of nonionic surfactants on modified DFPase efficiency and diffusional phenomena, was also examined. We will describe the stabilizing effects of the polymers in a future paper. Materials and Methods Materials. Hypol prepolymer and Pluronic surfactants, used in synthesizing protein-containing polymers, were purchased from Hampshire Chemical Co. (Lexington, MA) and from BASF (Parsippany, NJ). Diisopropylfluorophosphate (DFP), Bradford reagent, bovine serum albumin, optima grade hexane (99.9% pure), and buffer salts were purchased from Sigma-Aldrich Chemical Co. (St Louis, MO). DFPase was received as a kind gift from Dr. Ru¨terjans, Institut fu¨r Biophysikalische Chemie, Johann Wolfgang Goethe-Universita¨t, Frankfurt, Germany, after purification by Dr. Stefan Dierl. Protein-Containing Polymer Synthesis. DFPase-containing polyurethane foams were synthesized following the procedure we have published previously.12 All polymers were prepared using buffered aqueous mixtures (50 mM Bis-TrisPropane buffer, pH 7.5, 5 mM CaCl2). When studying the effects of salt on immobilized DFPase, the buffered medium was supplemented with NaCl (0.5 M). In the case of immobilization in the presence of surfactant, the surfactant concentration reached a level of 1% (w/w) in the aqueous media. The aqueous mixture (5 mL) was poured into a cylindrical vessel, and followed by the addition of enzyme (0.05-10 mg). Hypol 3000 (5 g), a prepolymer derived from the toluene diisocyanate, was added to the DFPase solution, and the biphasic mixture was agitated for 30 s with a custom designed mixing head attached to a 2500 rpm hand held drill. After synthesis was completed, the cylindrical foam was weighed and then allowed to dry for 14 h under ambient conditions and weighed again. After drying, the foam may contain a residual amount of water. Since all foams are prepared in a similar manner, we expect that the fluctuations in the residual amount of water are not significant. Activity of DFPase Polyurethanes. Soluble enzyme was assayed in a 10 mL reactor in the presence of DFP (3 mM). The buffered medium in use contained 5 mM CaCl2 and 50 mM Bis-Tris-Propane, pH 7.5. Some experiments were performed in a buffered solution supplemented with 0.5 M NaCl. As DFPase acts by binding and hydrolyzing DFP (see below), the activity was measured by following fluoride release with a fluoride ion electrode. Immobilized enzyme

was assayed in a similar manner, using blocks of DFPasefoam cut from bulk-synthesized sample and ranging in weight from 0.08 to 0.02 g. Typically, the cubes were then placed in 12 mL of 3 mM DFP buffered solution and agitated by

Drevon and Russell

magnetic stirring. Fluoride bulk solution concentration was measured every 20 s for 3 min. To validate the sensor approach, we also followed DFP consumption directly.18 In a 10 mL reactor containing 3 mM DFP solution, samples of 0.1 mL were taken over time and added to 0.1 mL of hexane. The biphasic mixtures were then equilibrated at room temperature with a constant shaking speed of 2500 rpm for 5 min and centrifuged to separate the organic and aqueous phases. The upper phase in each sample was analyzed with a Hewlett-Packard model 5890 series II gas chromatograph system (Hewlett-Packard, Wilmington, DE) equipped with a nitrogen-phosphorus detector (NDP). An Alltech EC-1 capillary column with a 30 m × 0.53 mm i.d., 1.2 µm film thickness, composed of a 100% poly(dimethylsiloxane) phase, was used. The temperature was programmed from a 50 °C initial temperature with a 1.5 min initial time to a 190 °C final temperature with a 7 min final time, applying a ramp rate of 20 °C/min and leading to a total elapsed run time of 15.5 min. The injector and detector temperatures were held constant at 250 and 300 °C, respectively. Runs were performed using 0.5 µL injection volumes, air and hydrogen gas flow rates of 100 and 3 mL/min, respectively, and a 21 pA baseline offset. Helium was used as both the carrier and makeup gas, with carrier and makeup gas flow rates of 11 and 19 mL/min, respectively. Product and Substrate Partitioning. The product and substrate partitioning between the foam and the bulk solution need to be evaluated in order to determine the activity retention associated with the immobilization process and the kinetic characteristics of DFPase-containing bioplastics. Since the use of surfactants during the Bioplastic synthesis affects the structure and properties of polyurethanes, it is necessary to assess the dependence of partition coefficients on foam formulations. The fluoride ion partitioning was estimated by following the enzymatic activity of soluble DFPase in the absence and the presence of foam blocks not containing enzyme. Additionally, various buffered solutions (50 mM Bis-TrisPropane, 5 mM CaCl2, pH 7.5) of fluoride ion at both high and low salt concentrations (0.5 M NaCl) were prepared and examined with a fluoride ion electrode. Blank foam samples were then inserted into the fluoride ion solutions, and after 20 min of magnetic stirring at room temperature, the fluoride ion concentrations were remeasured. To determine the substrate partition coefficients ([DFP]aqueous/[DFP]foam), blank foam blocks were added to buffered solutions containing various DFP concentrations, and the resulting biphasic systems were magnetically agitated for a 20-min period. Samples of the final solutions were then used to assay the activities of fixed soluble DFPase amounts. Knowing the kinetic constants of native DFPase, the residual DFP concentrations were calculated and compared to the substrate concentrations obtained by following the same procedure in absence of foam blocks. Determination of Kinetic Constants. The MichaelisMenten equation was applied as a kinetic model, and the kinetic constants for both soluble and immobilized DFPase were calculated using nonlinear regression analysis and the algorithm of Marquardt-Levenberg (SigmaPlot Version 2.0).

Immobilization of Diisopropylfluorophosphatase

Biomacromolecules, Vol. 1, No. 4, 2000 573

The observed reaction rates and substrate concentrations were corrected using the substrate and product partitioning coefficients measured as described above. Protein Concentration Determination. Protein concentrations were evaluated using the Bradford reagent.19 The addition of the dye to protein solutions at room temperature resulted in the formation of a dye-protein complex within 15 min, with an absorption maximum at 596 nm. A calibration curve with an extinction coefficient of 0.0341 µg-1 mL was obtained for protein concentrations ranging from 1 to 10 µg/mL. Results Reversibility of DFPase Attachment. The extent to which DFPase is irreversibly (covalently) attached to the polymer was determined using the Bradford reagent. Foam blocks cut from DFPase-containing polymers with a protein content of approximately 0.3% (by weight) were extensively rinsed with distilled water. Less than 1% (w/w) of the protein amount present within the blocks was detected in the rinsates, indicating that the immobilization efficiency approached 100%. Substrate and Product Partitioning. The enzymatic activity of soluble DFPase was not influenced by the presence of blank foams. The product (fluoride ions) partition coefficient was determined to be 1 and was independent of the type of polyurethane polymer formulation. DFP is hydrophobic enough to become concentrated in the foam, with an equilibrium partitioning coefficient of 1.3 (Table 1). In the presence of salt in the aqueous media, the use of surfactant L62 further accentuates this partition, increasing the partition coefficient to 1.5. Table 1. DFP Partitioning into Bioplasticsa surfactant salt salt no salt no salt

L62 L62

degree of swellability (DS)

DFP partitioning coefficient (PC)

5.9 10.5 6.2 11.4

1.3 ( 0.4 1.5 ( 0.3 1.3 ( 0.3 1.3 ( 0.2

a Details are provided in the text. DS ) m - m /m , where m is the w d d w mass of wet foam block and md the mass of dry foam block. P ) [DFP]a/ [DFP]f, where [DFP]a is the DFP concentration in the bulk solution and [DFP]f is the DFP concentration in the foam block.

Activity in Absence of Surfactant. To assess the enzyme distribution within the polyurethane, the DFPase-containing polymer was cut seVeral times in the axial direction, and assayed for enzymatic actiVity. The results imply that DFPase was homogeneously distributed in the foam since the actiVity did not fluctuate significantly between the regions. As shown in Figure 1, the effect of enzyme concentration on DFPase polyurethane activity is nonlinear, with saturation occurring above 0.5 mgDFPase/gfoam. Below approximately 0.1 mgDFPase/gfoam, the enzymatic activity is directly proportional to enzyme loading. To further investigate DFPase-polyurethane polymer properties, we determined the DFPase concentration dependence of apparent KM and kcat for polymers containing 0.02-0.64 mg/gfoam (Table 2, experiments 3 and 4). A 6-fold increase in enzyme loading provoked a

Figure 1. Effect of DFPase concentration on DFPase-containing polymer efficiency. Enzymatic activity measured using NPD-GC (opened circles) and fluoride sensor (closed circles) in buffered solutions (50 mM Bis-Tris-Propane, 5 mM CaCl2, 0.5 M NaCl) at pH 7.5. The activity of the bioplastic is reported at a 3 mM DFP concentration.

consequent decrease in apparent kcat, coupled with an increase in the apparent KM. Given the possible appearance of mass transfer limitations in polymers containing high concentrations of DFPase, we investigated the effect of particle size on activity and kinetics. The polyurethane blocks were ground into particles with a diameter ranging from 0.1 to 0.3 mm, and apparent KM and kcat were measured once again for particles containing 0.02-0.64 mg/gfoam. Experiments 5 and 6 (Table 2) demonstrate that for the small particles neither apparent KM nor kcat was sharply dependent on DFPase concentration in the polymer. The native enzyme has a turnover number (kcat) of 219 ( 5 s-1 and a KM of 0.53 ( 0.04 mM in buffered media (50 mM Bis-Tris-Propane, 5 mM CaCl2) containing a high salt concentration (0.5 M NaCl) at pH 7.5 at room temperature (Table 2, experiment 1). Activity with Surfactants. Bioplastics synthesized in the presence of nonionic surfactants L62, P65, and F68 display greater apparent activity retention than those prepared without surfactant (Figure 2). The use of surfactant L62 generates bioplastics with the highest levels of apparent activity retention, and thus subsequent experiments were performed with L62-containing-DFPase polyurethanes. Interestingly, the L62-containig bioplastics exhibited kinetic properties which were less enzyme loading dependent (Table 2, Experiments 7 and 8). Effect of Surfactant on Polymer Morphology. Our interest in the underlying reasons for why the activity of the L62-containing materials were less dependent on loading led us to study the morphological effects of the surfactant with electron microscopy (Figure 3). The addition of L62 results in an enlargement of the average micropore diameter, and the appearance of pores within the pores. Given the clear internal mass transfer effects controlling the DFPase poly-

574

Biomacromolecules, Vol. 1, No. 4, 2000

Drevon and Russell

Table 2. Kinetic Parameters for DFPase-Containing Polymers and Soluble DFPasea experiment 1b 2b 3 4 5c 6c 7 8 9 10 11 12c 13c

salt loading (M)

surfactant

0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

L62 L62 L62 L62 L62

agitation (v, rpm) magnetic magnetic magnetic magnetic magnetic magnetic magnetic magnetic mechanical (600) mechanical (800) mechanical (1000) magnetic magnetic

DFPase concn (mg/gfoam)

apparent KM (mM)

apparent kcat,app (s-1)

apparent kcat/KM (s-1 mM-1)

0.02 0.64 0.02 0.64 0.02 1.66 0.03 0.03 0.03 0.04 1.27

0.53 ( 0.04 0.79 ( 0.02 1.2 ( 0.1 2.9 ( 0.4 0.7 ( 0.1 1.1 ( 0.1 0.46 ( 0.04 1.6 ( 0.2 0.30 ( 0.06 0.34 ( 0.05 0.36 ( 0.04 1.6 ( 0.2 1.3 ( 0.2

219 ( 5 232 ( 2 49 ( 3 14 ( 1 46 ( 4 43 ( 2 113 ( 3 83 ( 3 87 ( 3 103 ( 4 94 ( 3 54 ( 3 45 ( 3

413 ( 41 294 ( 10 41 ( 6 5(1 66 ( 15 39 ( 5 249 ( 28 52 ( 8 290 ( 68 303 ( 56 261 ( 37 34 ( 6 35 ( 8

a The kinetic parameters were evaluated at room temperature in buffered media (50 mM Bis-Tris-Propane, 5 mM CaCl , pH 7.5) using substrate 2 concentrations varying from 0 to 20 mM, by applying the Michaelis-Menten equation as a model and using a nonlinear regression (Sigmat Plot Version b c 2). The errors on specific constants were calculated as follows: ∆(kcat/KM) ) (kcat/KM)[∆kcat/kcat + ∆KM/KM]. Native DFPase. Polyurethane blocks ground into particles.

4; Table 2, experiments 12 and 13). Once again, the addition of L62 during synthesis considerably reduces the diffusional limitations as shown by the increase in apparent activity. Discussion

Figure 2. Effect of surfactant on DFPase-polymer efficiency. Foams were synthesized without surfactant (closed circles), and containing P65 (closed squares), F68 (closed triangles), and L62 (closed diamond) in buffered solutions (50 mM Bis-Tris-Propane, 5 mM CaCl2, 0.5 M NaCl) at pH 7.5. The activity of the bioplastic is reported at a 3 mM DFP concentration.

urethane activity, we investigated the role of external mass transfer by measuring the agitation dependence of the kinetic constants (Table 2, experiments 9, 10 and 11). As expected, the KM and kcat values do not significantly vary with agitation speed. Moreover, they are similar to those obtained using magnetic stirring at a 0.02 mg/gfoam immobilized DFPase loading (Table 2, experiments 3, 4, and 7). Effect of Salt Removal on Enzyme Activity. Soluble DFPase promotes a turnover number of 232 ( 2 s-1 and a KM of 0.79 ( 0.02 in buffered media (50 mM Bis-TrisPropane, 5 mM CaCl2) at pH 7.5, under ambient conditions (Table 2, experiment 2). Given the known effect of salt on the activity of soluble DFPase, we investigated the effect of salt (0.5 M NaCl) on activity retention and transport effects in the biopolyurethane. The loading dependence of activity, and its elimination upon crushing, is not affected by the removal of salt (Figure

The saturation of bioplastic catalytic activity at high enzyme loading, combined with a significant increase in apparent KM and a decrease in apparent kcat, is a strong indication of the presence of diffusional limitations. Indeed, a 10-fold decrease in catalytic efficiency was observed for polyurethanes containing 0.02 and 0.64 mgDFPase/gfoam. At low DFPase loading the direct proportionality between apparent activity and enzyme concentration implied the absence of mass transfer limitations. Additionally, at low loading we observed that the activities for block foams and crushed foams were identical, and that no change in the kinetic constants was found by varying the agitation type and speed. Therefore, we again concluded that the activities of the polyurethanes at low enzyme loading were not diffusionally limited. Since the DFPase-containing polymers were under kinetic control at low loading, their intrinsic kinetic characteristics were evaluated. The retention of 67% activity under these conditions is characteristic of other multipoint immobilization techniques.1,5,9,12 At high enzyme loading, recovery of the activity predicted from the intrinsic kinetic data could only be achieved after crushing. Clearly, at these loadings the bulk bioplastics were limited by internal mass transfer. L62-containing polyurethanes displayed greater activity retention than those prepared without surfactant (6-fold increase in apparent intrinsic specificity constant), and in this case, an increase in enzyme loading from 0.02 and 1.66 mg/ gfoam resulted in only a 4-fold decrease in catalytic efficiency (the degree to which rate is reduced relative to that predicted from solution kinetics). Clearly, there are less severe diffusional limitations in the presence of the surfactant. Surface properties of supports for enzyme immobilization, such as porosity, have been previously investigated with scanning electron microscopy.20 Depending on the type of surfactant used, polyurethane foams display both macro- and microporosity.13,21 The use of L62 resulted in a significant

Immobilization of Diisopropylfluorophosphatase

Biomacromolecules, Vol. 1, No. 4, 2000 575

Figure 3. Photographs and electron micrographs of polymers prepared without surfactant (figures a and b) and with L62 (figures c and d) as described in the text.

synthesis using a foamable prepolymer (Hypol 3000). The results show that the activity of DFPase-containing bioplastics is limited by internal diffusion. The addition of nonionic Pluronic surfactants during the immobilization process changes the foam macro- and microstructure, leading to an enhancement of apparent and intrinsic catalytic efficiency. When synthesized with L62, the efficiency of DFPase-foam was 67% of that of the soluble enzyme. A buffered solution containing 1.05 × 10-4 wt % of immobilized DFPase is capable of achieving the same rate of DFP degradation as a 0.1 wt % calcium hypochlorite bleach. Given the significant stabilization that immobilization can be expected to provide, the resulting catalyst should be an effective decontaminant for a variety of nerve agents. The thermostability of natiVe and immobilized DFPase at Various temperatures will be presented in a future paper.

Figure 4. Effect of DFPase loading on DFPase-containing polyurethane efficiency in the absence of salt. Foams not containing surfactant (open triangles) and containing L62 (closed triangles) in buffered solutions (50 mM Bis-Tris-Propane, 5 mM CaCl2) at pH 7.5. The activity of the bioplastic is reported at a 3 mM DFP concentration.

enlargement of the micropores (Figure 3, parts a and c) and also changed the surface contours within the micropores (Figure 3, parts b and d). The increased micropore diameter may therefore explain the reduction in mass transfer limitation, by facilitating the substrate and product diffusion within the foam. Conclusion Covalent incorporation of DFPase into polyurethane foams has been performed in a single step protein-polymer

Acknowledgment. This work was funded by the German Ministry of Defense. We thank Dr. Ru¨terjans, Dr. S. Dierl and J. Hartleib for providing the enzyme. References and Notes (1) Lejeune, K. E.; Wild, J. R.; Russell A. J. Nerve agent degraded by enzymatic foams. Nature 1998, 395, 27-28. (2) Yang, Z.; Domach, M.; Auger, R.; Yang, F. X.; Russell, A. J. Poly(ethylene glycol)-induced stabilization of subtilisin. Enzyme Microb. Technol. 1996, 18, 82-89. (3) Wang, S. L.; Chio, S. H. Reversible immobilization of Chitinase via coupling to reversibly soluble polymer. Enzyme Microb. Technol. 1998, 22, 634-640. (4) Panza, J. L.; LeJeune, K. E.; Venkatasubramanian, S.; Russell, A. J. Incorporation of PEG-Proteins into polymers. In Poly(ethylene glycol) Chemistry and Biological Application; ACS Symposium Series 680; American Chemical Society: Washington, DC, 1997; Chapter 9. (5) Storey, K. B.; Duncan, J. A.; Chakrabarti, A. C. Immobilization of amyloglucosidase using two forms of polyurethane polymer. Appl. Biochem. Biotechnol. 1990, 23, 221-236.

576

Biomacromolecules, Vol. 1, No. 4, 2000

(6) Fulcrand, V.; Jacquier, R.; Lazaro, R.; Viallefont, P. Enzymatic peptide synthesis in organic solvent mediated by gels of copolymerized acrylic derivatives of R-chymotrypsin and polyoxyethylene. J. Peptide Protein Res. 1991, 38, 273-277. (7) Dias, S. F.; Vilas-Boas, L.; Cabral, J. M. S.; Fonseca, M. M. R. Production of ethyl butyrate by Candida rugosa lipase immobilized in polyurethane. Biocatalysis 1991, 5, 21-34. (8) Havens, P. L.; Rase, H. F. Reusable immobilized enzyme/polyurethane sponge for removal and detoxification of localized organophosphate pesticide spills. Ind. Eng. Chem. Res. 1993, 32, 22542258. (9) Wang, P.; Sergeeva, M. V.; Lim, L.; Dordick, J. S. Biocatalytic plastics as active and stable materials for biotransformations. Nature Biotechnol. 1997, 115, 789-793. (10) Yang, Z.; Williams, D.; Russell, A. J. Synthesis of proteins-containing polymers in organic solvents. Biotechnol. Bioeng. 1995, 45, 10-17. (11) Yang, Z.; Mesiano, A. J.; Venkatasubramanian, S.; Gross, S. H.; Harris, J. M.; Russell, A. J. Activity and stability of enzymes incorporated into acrylic polymers. J. Am. Chem. Soc. 1995, 117, 4843-4850. (12) LeJeune, K. E.; Russell, A. J. Covalent binding of a nerve agent hydrolyzing enzyme within polyurethane foams. Biotechnol. Bioeng. 1996, 51, 450-457. (13) LeJeune, K. E.; Mesiano, A. J.; Bower, S. B.; Grimsley, J. K.; Wild, J. R.; Russell, A. J. Dramatically stabilized phosphotriesterasepolymers for nerve agent degradation. Biotechnol. Bioeng. 1997, 54, 105-114.

Drevon and Russell (14) LeJeune, K. E.; Swers, J. S.; Hetro, A. D.; Donahey, G. P.; Russell, A. J. Increasing the tolerance of organophosphorus hydrolase to bleach. Biotechnol. Bioeng. 1999, 64, 250-254. (15) Braatz, J. A. Biocompatible polyurethane-based hydrogel. J. Biomater. Appl. 1994, 9, 71-96. (16) Hoskin, F. C. G. Diispropylphosphorofluoridate and tabun: enzymatic hydrolysis and nerve function. Science 1971, 172, 1243-1245. (17) Hoskin, F. C. G.; Prusch, R. D. Characterization of a DFP-hydrolyzing enzyme in squid posterior salivary gland by use of soman, DFP and manganous ion. Comp. Biochem. Physiol. 1983, 75C, 17-20. (18) Czerwinski, S. E.; Maxwell, D. M.; Lenz, D. E. A method for measuring octanol:water partition coefficients of highly toxic organophosphorus compounds. Toxicol. Methods 1998, 8, 139-149. (19) Sedmak, J. J.; Grossberg, S. E. A rapid method, sensitive and versatile assay for protein using coomassie brilliant blue G250. Anal. Biochem. 1977, 79, 544-552. (20) Liu, Y.; Qian, J.; Fu, X.; Liu, H.; Deng, J.; Yu, T. Immobilization of horseradish peroxidase onto a composite membrane of regenerated silk fibroin and poly(vinyl alcohol) and its application to a new methylene blue-mediating sensor for hydrogen peroxide. Enzyme Microb. Technol. 1997, 21, 154-159. (21) Hu, Z.; Korus, R. A.; Stormo, K. E. Characterization of immobilized enzymes in polyurethane foams in a dynamic bed reactor. Appl. Microbiol. Biotechnol. 1993, 39, 289-295.

BM0000034