Mechanistic and Molecular Investigations on Stabilization of

Anal. Chem. 2002, 74, 3037-3045

Mechanistic and Molecular Investigations on Stabilization of Horseradish Peroxidase C Anja Schmidt,† Jens T. Schumacher,‡ Joachim Reichelt,† Hans-Juergen Hecht,† and Ursula Bilitewski*,‡

Department of Molecular Structural Research and Division of Biochemical Engineering, National Research Centre for Biotechnology Ltd., Mascheroder Weg 1, D-38124 Braunschweig, Germany

The enzyme horseradish peroxidase (HRP) shows a decreasing activity when the enzyme’s substrate hydrogen peroxide is present with the degree of inactivation being dependent on the incubation time and the hydrogen peroxide concentration. Incubation times of some minutes do not inactivate the enzyme independent of the H2O2 concentration. After several hours, only 50% of the activity is found for a medium H2O2 excess, and a >100-fold excess of H2O2 completely inactivates the enzyme. Polymeric additives, in particular Gafquat, lead to higher residual activities, whereas stabilizers, such as aminopyrine, preserve the full activity. Circular dichroism (CD) measurements reveal that the enzyme structure remains more or less unchanged when hydrogen peroxide is added. Only when a 1000-fold excess of hydrogen peroxide is present are structural changes observed. UV spectra highlight that the heme group in the enzyme is affected by hydrogen peroxide in a first step. Without any prolonged incubation, a decrease of the Soret band to ∼50% is found for low hydrogen peroxide concentrations (HRP/H2O2 from 1:1 to 1:100). Higher H2O2 concentrations lead to the formation of catalytically inactive HRP forms. Preincubation of Gafquat, which is a copolymer from vinylpyrrolidone and derivatized methyl methacrylate, with hydrogen peroxide shifts the influence of hydrogen peroxide to higher concentrations, the shift being dependent on the Gafquat concentration. This effect is not observed for other polymers, such as dextrans, but it is also found for the stabilizer aminopyrine. Extended incubation times (24 h) of HRP together with H2O2, however, lead to an at least partial recovery of the Soret band for lower H2O2 concentrations (H2O2/HRP from 1:1 to 1:100). When hydrogen peroxide is used in a >100 fold excess, the heme group is irreversibly destroyed, and even the characteristic band of cpd III is not found. Here, the presence of Gafquat only reduces the degree of destruction. Computer modeling of the interaction between the polymers and the enzyme shows no specific binding sites for the functional groups of the vinylpyrrolidone-methacrylate copolymer Gafquat or of DEAEdextran on the enzyme, whereas for the only activating † ‡

Department of Molecular Structural Research. Division of Biochemical Engineering.

10.1021/ac0108111 CCC: $22.00 Published on Web 05/22/2002

© 2002 American Chemical Society

polymer, polyethylenimine clustering of binding sites is observed. Horseradish peroxidase (HRP) is widely used as indicator enzyme in enzyme immunoassays,1 enzyme electrodes,2,3 and nucleic acid detection systems.4 It is a donor/hydrogen peroxide oxidoreductase (EC 1.11.1.7) and catalyses the oxidation of a variety of organic and inorganic substances with H2O2 as electron acceptor.5-8 Its isoenzyme C is the most studied member of the heme-containing plant peroxidases, the crystal structure of which was recently elucidated by Gajhede et al.9 The catalytic cycle of HRP involves the oxidation and reduction of the Fe(III) porphyrin group and can be described by the following equations:5,7,10

HRP + H2O2 f cpd I + H2O

(1)

cpd I + AH2 f cpd II + AH*

(2)

cpd II + AH2 f HRP + AH*

(3)

Native HRP (ferriperoxidase: FeIII, Por) is oxidized by H2O2 through a formally two-electron-transfer reaction to compound I (cpd I), in which the heme group is oxidized to an unstable oxyferryl radical intermediate (FeIVdO, Por•+) with a formal oxidation state of 5+.11 The enzyme is regenerated in two oneelectron-transfer reactions by an electron donor (AH2) (eqs 2, 3), delivering a nonradical intermediate, compound II (cpd II) (1) Thijssen, P. Practice and Theory of Enzyme Immunoassay; Elsevier Science Publishers B. V.: Amsterdam, 1985; pp 173-219. (2) Ruzgas, T.; Cso¨regi, E.; Emneus, J.; Gorton, L.; Marko-Varga, G. Anal. Chim. Acta 1996, 330, 123-138. (3) Popescu, I. C.; Zetterberg, G.; Gorton, L. Biosens. Bioelectron. 1995, 10, 443-461. (4) Pollard-Knight, D.; Read, C. A.; Dowes, M. J.; Howard, L. A.; Leadbetter, R. R.; Pheby, S. A.; McNaughton, E.; Syms A.; Brady, M. A. W. Anal. Biochem. 1990, 185, 84-90. (5) Yamazaki, I. In Molecular Mechanisms of Oxygen Activation; Hayaishi, O., Ed.; Academic Press: New York, 1974; Chapter 13, pp 535-558. (6) Dunford, H. B.; Stillman, J. S. Coord. Chem. Rev. 1976, 19, 187-251. (7) Dunford, H. B. In Peroxidases in Chemistry and Biology; Everse, J., Everse, K. E., Grisham, M. B., Eds.; CRC Press: Boca Raton, 1991; Vol. 2, Chapter 1, pp 1-24. (8) Zhu, M.; Huang, X.; Shen, H. Talanta 2001, 53, 927-935. (9) Gajhede, M.; Schuller, D. J.; Henriksen, A.; Smith, A. T.; Poulos, T. L. Nat. Struct. Biol. 1997, 4, 1032-1038. (10) Saunders, B. C.; Holmes-Siedle, A. G.; Stark, B. P. In Peroxidase; Butterworth: London, 1964. (11) Adediran, S. A. Arch. Biochem. Biophys. 1996, 327, 279-284.

Analytical Chemistry, Vol. 74, No. 13, July 1, 2002 3037

(FeIVdO, Por) (eq 2) with an oxidation state of 4+, and the free radical AH*. In the absence of reducing substances, AH2, H2O2 can perform a dual role as oxidant in the formation of cpd I and as a typical one-electron donor substrate delivering the superoxide anion.11-13 The electron transfer reactions mentioned above are accompanied by proton transfer reactions, and thus, the reaction sequence represents only the simplified form usually used for the description of the peroxidase cycle. In combination with certain electron donors, the reduction of cpd I to the native enzyme can also proceed in a single step. This reaction is also observed if only H2O2 is present in a sufficient excess. Under these conditions, HRP shows a catalase-like activity producing oxygen through the oxidation of H2O2 by cpd I.13 Another form of HRP is known as compound III (cpd III).11-15 Cpd III is formed via cpd II with an excess of H2O2 and is catalytically not active. However, it is known that cpd III slowly decomposes to the native enzyme, thus leading to the restoration of the enzyme activity.13-15 An irreversibly inactive compound of HRP is slowly formed from cpd III11,12 or from cpd I13 when the excess of H2O2 is permanent, the so-called P670 with an absorbance maximum at 670 nm, which is also called cpd IV.14,16 The formation of P670 is the major inactivation process in which H2O2 acts as a suicide substrate. The mechanisms are still not completely understood.17-22 Analysis of the kinetics of the inactivation reactions and investigations with respect to its prevention revealed that the presence of electron donors AH2 influence the formation of inactive HRP forms.14,16,23 Depending on the chosen electron donor, the amount of hydrogen peroxide, and the concentration ratio of H2O2 and the electron donor, the enzyme is protected or inactivated. The inactive cpd III and cpd IV are formed from the reaction of H2O2 with the intermediate cpd I. Reactivation of cpd III also occurs, at least partly, via a reaction of cpd III with cpd I and cpd II.24 Thus, the electron donor competes with H2O2 and cpd III for cpd I, with the reaction with H2O2 leading to inactivation via P670 (cpd IV), the slow reaction with cpd III and the fast reaction with the electron donor to regeneration (eq 2) of the enzyme. A low ratio of H2O2 to electron donor protects the enzyme if the radicals (12) Arnao, M. B.; Acosta, M.; del Rio, J. A.; Varoń, R.; Garcia-Cańovas, F. Biochim. Biophys. Acta 1990, 1041, 43-47. (13) Hernandez-Ruiz, J.; Arnao, M. B.; Hiner, A. N. P.; Garcia-Canovas, F.; Acosta, M. Biochem. J. 2001, 354, 107-114. (14) Baynton, K. J.; Bewtra, J. K.; Biswas, N.; Taylor, K. E. Biochim. Biophys. Acta 1994, 1206, 272-278. (15) Adediran, S. A.; Lambeir, A.-M. Eur. J. Biochem. 1989, 186, 571-576. (16) Arnao, M. B.; Acosta, M.; del Rio, J. A.; Garcia-Cańovas, F. Biochim. Biophys. Acta 1990, 1038, 85-89. (17) Acosta, M.; Arnao, M. B.; del Rio, J. A.; Garcia-Cańovas, F. In Biochemical, Molecular and Physiological Aspects of Plant Peroxidases; Lobarzewski, J., Greppin, H., Penel, C., Gaspar, T., Eds.; University of Geneva: Geneva, 1991; pp 175-184. (18) Acosta, M.; Arnao, M. B.; Hernandez-Ruiz, J.; Garcia-Cańovas, F. In Plant Peroxidases: Biochemistry and Physiology; Welinder, K. G., Rasmussen, S. K., Penel, C., Greppin, H., Eds.; University of Geneva: Geneva, 1993; pp 201-205. (19) Hiner, A. N. P.; Hernandez-Ruiz, J.; Garcia-Cańovas, F.; Smith, A. T.; Arnao, M. B.; Acosta, M. Eur. J. Biochem. 1995, 234, 506-512. (20) Moosavi Movahedi, A. A.; Nazari, K.; Ghadermarzi, M. Ital. J. Biochem. 1999, 48, 9-17. (21) Arnao, M. B.; Garcia-Cańovas, F.; Acosta, M. Biochem. Mol. Biol. Int. 1996, 39, 97-107. (22) Nakajima, R.; Yamazaki, I. J. Biol. Chem. 1980, 255, 2067-2071. (23) Schu ¨ tz, A. J.; Winklmair, M.; Weller, M. G.; Niessner, R. SPIE 1997, 3105, 332-340. (24) Tamura, M.; Yamazaki, I. J. Biochem. 1972, 71, 311-319

3038 Analytical Chemistry, Vol. 74, No. 13, July 1, 2002

Figure 1. Chemical structues of the polymer Gafquat 755N (a), polyethyleneimine (b), and the “monomer” aminopyrine (c).

AH* (see eq 2, 3) are not too reactive.16 Moreover, it was observed that some of the cosubstrates stabilize HRP as additives during storage even when no H2O2 was present.23,25-27 This was attributed at least partly to the radical scavenging activity of those compounds that may be formed when the enzyme is not protected from light.23 The maintenance of the activity of enzymes in use and during storage, that is, their stability, is an essential parameter for all applications. Therefore, there are numerous investigations of additives to enhance protein stability under various conditions. With respect to an improved storage stability, the processes related to freezing and drying of proteins28 are of importance, but thermostability29 and stability of proteins in solution30 are also investigated. Because the maintenance of the 3D structure of proteins is a prerequisite for their activity, a low stability is often attributed to protein unfolding due to the flexibility of the amino (25) Gallati, H.; Brodbeck, H. European Patent Office EP 0 070 992 A1, 1982. (26) Klenner, D.; Kleinhammer, G.; Deeg, R. Deutsches Patentamt DE 3511327 A1, 1985. (27) Wehner, R.; Mattersberger, J.; Klenner, D. Deutsches Patentamt DE 3509238 A1, 1985 (28) Carpenter, J. F.; Izutsu, K.-I.; Randolph, T. W. Drugs Pharm. Sci. 1999, 96, 123-160. (29) Kumar, S.; Tsai, C.-J.; Nussinov, R. Protein Eng. 2000, 13, 179-191. (30) Jaenicke, R. J. Biotechnol. 2000, 79, 193-203.

Figure 2. Influence of H2O2 and Gafquat 755N on the activity of HRP during storage: (a) residual activity, depending on the incubation time; (b) residual activity, depending on the H2O2 concentration.

acid chain. Thus, the stabilizing effect of immobilization procedures21 and of additives, such as sugars,32 sugar alcohols, and polymers,33-36 is explained by the reduced flexibility of the protein due to (specific) interactions between the protein and the additives or modified interactions between protein and solvent, namely (31) Schellenberger, A.; Ulbrich, R. Biomed. Biochim. Acta 1989, 48, 63-67. (32) Wimmer, R.; Olsson, M.; Petersen, M. T. N.; Hatti-Kaul, R.; Petersen, S. B.; Mu ¨ ller, N. J. Biotechn. 1997, 55, 85-100. (33) Schumacher, J. T.; Mu ¨ nch, I.; Richter, T.; Rohm, I.; Bilitewski, U. J. Mol. Catal. B: Enzym. 1999, 7, 67-76. (34) Gavalas, V. G.; Chaniotakis, N. A.; Gibson, T. D. Biosens. Bioelectron. 1998, 13, 1205-1211. (35) Gibson, T. D.; Pierce, B. L. J.; Hulbert, J. N.; Gillespie, S. Sens. Actors B 1996, 33, 13-18. (36) Gibson, T. D.; Hulbert, J. N.; Woodward, J. R. Anal. Chim. Acta 1993, 279, 185-192.

water. Thus, it is believed that the stabilizing effects of the cationic polymers diethylaminoethyl (DEAE) dextran and Gafquat 755N (a copolymer of vinylpyrollidone and quaternized methyl methacrylate) observed for a number of enzymes are due to electrostatic protein-polyelectrolyte complexes.34-36 Most of the investigations of the improvement of protein stability by polymeric additives are aimed at improvements of the storage stability. Only in some applications is the operational stability considered.33-35,37 On the other hand, it is well-known that in particular, the enzyme HRP is inactivated by its own substrate H2O2 and that the inactivation is prevented, that is, the stability enhanced, by some low-molecular-weight additives (see above). (37) Rohm, I.; Genrich, M.; Collier, W.; Bilitewski, U. Analyst 1996, 121, 877881.

Analytical Chemistry, Vol. 74, No. 13, July 1, 2002

3039

Thus, we wanted to investigate the influence of polymers on the activity of HRP in the presence of its substrate H2O2. In particular, we wanted to know whether the stabilizing effect of polymers is to be attributed to the preservation of the 3D structure of the protein or to another effect. We chose the polymers Gafquat 755N (Figure 1a), dextran, dextran sulfate, and DEAE-dextran as examples, because they had proven to stabilize the activity of dissolved and immobilized HRP.33,34,36 The activity measurements were accompanied by molecular modeling of the interactions of polymer subunits with the protein and by analysis of the circular dichroism (CD) and UV spectra of the protein, leading to information on structural changes and effects on the heme group, respectively. EXPERIMENTAL SECTION Chemicals and Solutions. UV/vis spectra and CD experiments were performed in 66 mM sodium/potassium phosphate buffer pH 5.5 (NaKP, pH 5.5). The pH of a 66 mM KH2PO4 solution was adjusted with a 66 mM Na2HPO4 solution. Acetate buffer pH 5.5 was used for activity tests. The pH of a 100 mM sodium acetate solution was adjusted with 1% citric acid solution. The HRP substrate solution contained 100 µL of 1% H2O2 and 400 µL of 3,3′,5,5′-tetramethylbenzidine (TMB, Boehringer, Mannheim, Germany) solution (6 mg TMB in 1 mL of dimethyl sulfoxide (Merck-Schuchardt, Hohenbrunn, Germany)) in 25 mL of acetate buffer, pH 5.5. A 27.8 µM (1 mg/mL) HRP (isoenzyme C, Biozyme Laboratories Ltd., Gwent, U.K.) stock solution prepared in NaKP, pH 5.5, and kept at 4 °C in the dark was found to be stable for several weeks. Stock solutions of H2O2 (100 mM or 1%) were prepared in the required buffer and kept at 4 °C and were used for 2 days. Gafquat 755N (ISP, Guildford, UK), DEAEdextran (Sigma, Deisenhofen, Germany), dextran (Sigma), and dextran sulfate (Sigma) were used as additives. Aminopyrine (AP, 4-dimethylaminoantipyrine) was purchased from Aldrich (Steinheim, Germany). All other reagents were of analytical grade, and deionized water was used for preparation of solutions. Circular Dichroism (CD) Measurements. CD experiments were carried out using a Jasco J 600 spectropolarimeter (Japan Spectroscopic Co., Tokyo, Japan). CD spectra were recorded at room temperature in the presence of nitrogen in a wavelength range from 260 to 180 nm in cuvettes of 0.5-mm path length and a HRP concentration of 2.8 µM (0.1 mg/mL). The CD data were expressed in mean residue ellipticity values [Θλ, deg × cm2 × dmol-1]. The percentage of helix content was quantified using the computer program VARSELEC.38 UV/vis Spectra. UV/vis spectra of HRP were recorded on a UV/visible spectrophotometer (Ultrospec 3000, Pharmacia Biotech, Cambridge, U.K.). UV/vis spectra were monitored in the range from 280 to 700 nm vs a buffer solution containing H2O2 and the additive under investigation. The final HRP concentration was adjusted to 9.3 µM (0.33 mg/mL). H2O2 concentrations were varied from 1 µM to 15 mM H2O2 (final concentration). The incubation was started by adding an aliquot of the H2O2 stock solution to the HRP solution. Additionally, different compounds (Gafquat 755N, DEAE-dextran, dextran, dextran sulfate) in various concentrations were added to HRP prior to the addition of H2O2. The spectra were recorded immediately and 24 h after mixing. (38) Compton, L. A.; Johnson, W. C., Jr. Anal. Biochem. 1986, 155, 155-167.


Figure 3. Influence of the Gafquat concentration on the residual activity of HRP incubated for 24 h with H2O2.

Influence of H2O2 on HRP Activity. H2O2 with final concentrations from 10 µM to 15 mM was added to a HRP solution (9.3 µM). After specified time intervals, the remaining enzymatic activity was determined. The HRP/H2O2 solution was diluted 1:400 000, and 250 µL was pipetted into reaction tubes. A 500-µL portion of H2O2/TMB substrate solution was added. A blue intermediate complex was formed as a result of the reaction of HRP with H2O2 and TMB. A 150-µL portion of this reaction mixture was pipetted into microtiter plates (MaxiSorp, Nunc, Denmark). The activity of HRP was measured in microtiter plates to allow the parallel analysis of a number of samples and to reduce the amount of chemicals needed. After 10 min, the reaction was stopped by adding 50 µL of 1 M H2SO4. This turned the blue complex into a yellow dye, which was measured photometrically at 450 nm vs 650 nm.39 Each mixture was measured 3 times, and the mean value was taken. Additionally, the influences of Gafquat 755N and aminopyrine (10.4 mM) were tested by mixing with HRP prior to the addition of H2O2. Modeling. The potential interactions of stabilizer molecules with HRP were examined by docking calculations using the autodock program.40 The software determines preferential binding sites for a ligandsin this case, a subunit of the polymer (Figure 1a, 1b)son the receptor protein. For modeling the interaction of Gafquat’s two types of subunit was considered: n ) m ) 1 and n ) 2, m ) 1. In the polyethylenimine subunit a positive charge was associated with each nitrogen atom. Starting from random orientations of the ligand with respect to the protein, the affinities (hydrophobic, hydrophilic, and electrostatic) between receptor and ligand were calculated, and the ligand orientation was optimized for increased affinity. Flexibility of the ligand was taken into account by optimization of torsional angles, but the receptor molecule was regarded as rigid. For these calculations coordinates of glycosylated native HRP were modeled on the basis of the structure of recombinant nonglycosylated HRP (pdb-id 1ATJ). Structure models of the major (39) Josephy, P. D.; Eling, T.; Mason, R. P. J. Biol. Chem. 1982, 257, 36693675. (40) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.; Olson, A. J. J. Comput. Chem. 1998, 19, 1639-1662.

glycan species, identified by Gray et al.,41 were generated using the SWEET carbohydrate server42 and attached to the known glycosylation sites41 on the HRP structure. The carbohydrate structures attached to the protein were subjected to 70-ps molecular dynamics calculations43 carried out in a periodic boundary box filled with water for simulation of glycosylated HRP in solution. Partial atomic charges were assigned according to the AMBER43 force field. The polymer subunits, consisting of a short polymer backbone and two to three side chains, were constructed using Cerius2 3.0 (Molecular Simulations Inc.). Optimization of the structure and calculation of partial atomic charges were carried out using HyperChem 4.5 (HyperCube Inc.). The autodock40 runs consisted of 100 calculations on a cubic grid with lengths of 100 Å and a grid spacing of 1.8 Å, starting from random orientations and conformations of the ligand. Optimization was according to the Lamarckian genetic algorithm with population size 200 for a total of 15 000 000 energy calculations. Figures were generated using Molscript44 and Gras,45 and were rendered using gl_render (L. Esser, unpublished) and PovRay (http://www.povray.org). RESULTS Stability of Horseradish Peroxidase. Solutions of horseradish peroxidase with protein concentrations in the 100 µg/mL range could be stored at 4 °C for at least 6 months (data not shown). Dilution of the enzyme solution led to a decrease in the enzyme activity with time: solutions with 0.5 µg/mL HRP lost almost 50% of their activity within 14 days;33 and more dilute solutions (range ng/mL) had to be used immediately after dilution. However, if the enzyme’s substrate hydrogen peroxide (H2O2) was added, the activity of even concentrated enzyme solutions decreased significantly within 1 day, depending on the concentration of H2O2 (Figure 2a). Short incubation times, that is, ∼30 min., had almost no influence on the activity, irrespective of the concentration (Figure 2b). Extending the incubation to more than 24 h did not influence the activity further (data not shown). This correlates with the observation of Hernandez-Ruiz et al.13 that after 4 h, the endpoint is reached as a result of either total H2O2 consumption or total enzyme inactivation. In previous investigations, the addition of polymers, such as DEAE-dextran and Gafquat 755N (Figure 1), increased the stability of HRP solutions of medium concentrations.33 Thus, we investigated the influence of Gafquat 755N on the HRP activity in the presence of hydrogen peroxide. As can be seen in Figure 2, Gafquat 755N could not prevent the inactivation, but all enzyme solutions containing Gafquat 755N had a higher remaining activity than the corresponding solutions without Gafquat 755N. This effect depended on the Gafquat concentration (Figure 3). After (41) Gray, J. S.; Yang, B. Y.; Montgomery, R. Carbohydr. Res. 1998, 311, 6169. (42) Bohne, A.; Lang, E.; von der Lieth, C. W. J. Mol. Model. 1998, 4, 33-43. (43) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E., III; Ross, W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K. M.; Stanton, R. V.; Cheng, A. L.; Vincent, J. J.; Crowley, M.; Ferguson, D. M.; Radmer, R. J.; Seibel, G. L.; Singh, U. C.; Weiner, P. K.; Kollman, P. A. AMBER 5; University of California: San Francisco, 1997; http://www.amber.ucsf.edu/amber/index.html. (44) Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 946-950. (45) Nicholls, A.; Sharp, K. A.; Honig, B. Proteins 1991, 11, 281-296.

Figure 4. Stabilizing effect of aminopyrine (AP) on HRP stored for the specified time in the presence of H2O2.

Figure 5. CD spectra of HRP (a), in the presence of 5 mM H2O2 (b), and in the presence of 5 mM H2O2 and 0.2% Gafquat (c).

24 h incubation with H2O2, the residual HRP activity increased with the Gafquat concentration. Klenner et al.26 used aminopyrine (AP, Figure 1c) to stabilize the activity of HRP in solution and, thus, to enhance its storage stability. Aminopyrine is chemically similar to aminoantipyrine, a substrate of HRP, which also has a stabilizing effect on HRP.25 We found that AP (10.4 mM) was able to stabilize the HRP activity in the presence of H2O2 (0.5-10 mM) (Figure 4). Solutions were stored at 4 °C, and after 7 days the activity of HRP still remained. In contrast, solutions without AP lost in the same time almost 50% of their activity (0.5 mM H2O2), and with 10 mM H2O2, the remaining activity was negligible after 24 h. For a more detailed understanding of these effects, the possible effects of Gafquat and H2O2 on the enzyme structure were investigated with CD measurements and on the heme group with UV/vis spectral analysis. CD Spectra. The effect of H2O2 addition and the influence of Gafquat on the CD spectrum of HRP in the UV region were examined (Figure 5). Two negative bands at ∼222 and 208 nm and an intense positive band at 192 nm are characteristics of the CD spectra due to R-helical secondary structure elements.46,47 CD (46) Johnson, W. C., Jr. Proteins 1990, 7, 205-214.


3041

spectra analysis has already been successfully used to characterize the influence of temperature and pH on the secondary structure of HRP.48 The R-helix content decreased with increasing H2O2 content, as indicated by the decrease in ellipticity at 222 and 208 nm. Figure 5 shows the CD spectra of HRP without any admixture (trace a), in the presence of 5 mM H2O2 (trace b) and with 5 mM H2O2 plus 0.2% Gafquat (trace c). The spectra obtained in the presence of H2O2 showed a significant decrease in ellipticity, which was reduced in the additional presence of Gafquat. The addition of only Gafquat to HRP had no influence on the CD spectrum (data not shown). The corresponding R-helix contents were estimated to be 38% (a), 22% (b), and 31% (c). These results indicated that the exposure of HRP to H2O2 influenced the secondary structure of HRP, and this effect was reduced in the presence of Gafquat. It should be noted that with H2O2 concentrations lower than 5 mM (1000-fold excess to HRP), no significant change in ellipticity could be monitored (data not shown), whereas the decrease of the HRP activity was observed even in the presence of an only 10 µM H2O2 solution (low H2O2/HRP ratio). UV/vis Spectral Analysis. During the catalytic cycle (eqs 1-3), HRP is present as native enzyme, cpd I and cpd II. These 3 forms differ in the oxidation state of Fe in the center of the heme group and can be distinguished by characteristic absorption bands in the UV/vis spectra. Native HRP shows a strong maximum at 401.5 nm (Soret band) and much lower peaks at ∼500 and 640 nm. Cpd I has absorption bands at 410 and 657 nm, whereas maxima at 418, 527, and 558 nm are characteristic for cpd II. As mentioned earlier, inactive enzyme forms are known. Bands at 416, 546, and 583 nm indicate the presence of cpd III, and the formation of P670 is indicated by a band at 670 nm.10-12 These characteristics could be confirmed when the influences of H2O2 and Gafquat on HRP were investigated by UV/vis spectra (Figure 6). In the UV spectrum of native HRP, the strong absorption at ∼400 nm, the Soret band, was clearly identified (data not shown). Upon the addition of H2O2, the spectrum changed, with the extent of change being dependent on the peroxide concentration. As summarized in Figure 6a, the intensity of the Soret band decreased significantly when the H2O2 concentration exceeded 10 µM, and measurements were performed immediately after mixing of the components. The absorption band became not only smaller, but also wider with a kind of shoulder in the lower wavelength range (down to ∼350 nm). When the H2O2 concentration was increased to more than 1 mM (100-fold the HRP concentration), only cpd III could be identified through the main absorption band at 416 nm and smaller bands at 544 and 576 nm (∼10% of the main band). In addition, a small band at 670 nm was found, indicating the presence of P670. At H2O2 concentrations exceeding 5 mM, the intensity of the band at 416 nm reached the original intensity of the Soret band. These results were similar to those reported by George.49 It was reported that cpd III is regenerated to native HRP.11-13,24 This was confirmed by UV spectra obtained after an incubation of HRP with H2O2 for more than 24 h (Figure 6b). The initial intensity of the Soret band was almost restored for low to medium H2O2 concentrations. Only when the H2O2 concentration was higher than 1 mM, was the (47) Strickland, E. H.; Kay, E.; Shannon, L. M.; Horwitz, J. J. Biol. Chem. 1968, 243, 3560-3565. (48) Chattopadhyay, K.; Mazumdar, S. Biochemistry 2000, 39, 263-270. (49) George, P. J. Biol. Chem. 1953, 201, 427-434.


Figure 6. Intensity of the Soret band (401 nm) in the UV spectra of HRP (9.3 µM) incubated with H2O2. The intensity of the Soret band of HRP without H2O2 was taken as 100%. The mean of the OD401 was 0.68 (0.03. (a) Influence of Gafquat and the preincubation of Gafquat with H2O2. Measurements were performed directly after mixing; (b) measurements were performed 24 h after mixing; (c) influence of the polymer and the substrate analogue aminopyrine. All polymers were preincubated with H2O2 for 24 h. Measurements were performed immediately after mixing with HRP.

heme group almost totally destroyed, because the Soret band was significantly smaller, and characteristic bands of cpd III were not

Figure 7. Superposition of Gafquat subunits (ball and stick representation) from 100 docking calculations on HRP (semitransparent surface with CR trace and ball-stick representation for carbohdrate and heme group). Nitrogen atoms are coloured blue; oxygen atoms, red; and carbon atoms, black (Gafquat) and light brown (protein). The view is to the active site in the middle of the molecule.

found. The intensity of the band at 670 nm, indicating the presence of irreversibly inactive P670, was increased 3-fold compared to the previous measurements. The presence of Gafquat resulted in higher intensities of the Soret band for all H2O2 concentrations when measurements were done directly after HRP addition (Figure 6a). To investigate the stabilizing mechanism further, Gafquat and H2O2 were preincubated for 24 h prior to the addition of HRP. This led to a shift of the changes in the HRP spectrum to much higher H2O2 concentrations (Figure 6a). Using 0.66% Gafquat, no influence on the Soret band could be noted up to 1 mM H2O2. Higher H2O2 concentrations caused the shift of the absorption band to 416 nm, indicating the formation of cpd III. Decreasing the amount of Gafquat allowed the preincubation only with lower H2O2 concentrations without visible changes in the HRP spectra. Preincubation of DEAE-dextran, dextran, and dextran sulfate (each 0.66%) with H2O2 had much less influence on the reaction of HRP and H2O2, whereas the presence of aminopyrine showed the same effects as the presence of Gafquat (0.66%) (Figure 6c). Incubation of HRP in the polymer-H2O2 mixture for 24 h again showed the different behaviors of Gafquat and DEAE-dextran (Figure 6b): when H2O2 was present in concentrations from 10 µM to 1 mM (HRP/H2O2 from 1:1 to 1:100), the intensity of the Soret band was found to be ∼85% of the original intensity. Even in this range, preincubation of Gafquat with H2O2 had slightly beneficial effects, because the 3 upper curves originate from solutions with the preincubated polymer. In the higher H2O2 concentration range, the intensity of

the Soret band decreased with increasing H2O2 concentration; however, the decrease was more pronounced for the pure enzyme and the enzyme mixed with DEAE-dextran. Computer Modeling. Modeling the interactions of subunits of polymeric stabilizers with HRP showed Gafquat (Figure 1a) attached at numerous sites on the protein in various conformations (Figure 7), suggesting unspecific binding. Calculations with subunits of DEAE-dextran produced a similar result (data not shown), whereas subunits of polyethylenimine (Figure 1b), in contrast, clustered at a few negatively charged sites of HRP (Figure 8), indicating more specific electrostatic binding. Polyethylenimine was found to increase the initial activity of HRP enzyme electrodes, but no stabilizing effects were observed.33 Considering that only subunits of the polymers were used for modeling, which are connected to each other in the “real” polymer, gives the impression that the polymers Gafquat and DEAEdextran cover the protein surface more or less homogeneously, forming perhaps a kind of cage in which the protein is trapped and which may prevent enzyme unfolding. This picture is also compatible with the formation of complexes between proteins and polyelectrolytes.34,35 In contrast, the coverage of the protein by polyethylenimine is not as complete. DISCUSSION Addition of H2O2 is known to inactivate even concentrated HRP solutions, depending on concentration and incubation time, with the maximum degree of inactivation being obtained after several Analytical Chemistry, Vol. 74, No. 13, July 1, 2002

3043

Figure 8. Superposition of polyethylenimine subunits (ball and stick representation) from 100 docking calculations on HRP (semitransparent surface with CR trace and ball-stick representation for carbohydrate and heme group). Nitrogen atoms are coloured blue and carbon atoms black (polyethylenimine) and light brown (protein); calcium atoms are shown as brown spheres. The view is identical to Figure 7.

hours incubation. This effect was reduced in the presence of Gafquat and totally prevented by aminopyrine. CD spectra showed that this inactivation is related to structural changes only at higher H2O2 concentrations, and UV spectral analysis highlighted that in a first step, mainly the heme group is attacked. The Soret band, characteristic for the native heme group, is modified at once after H2O2 addition, even with only equimolar concentrations of HRP and H2O2. A 100-fold excess of H2O2 led to the accumulation of cpd III visible in the UV spectra, which is an inactive HRP form, but can react back to native HRP. Activity determinations for the investigation of the enzyme’s stability require the addition of electron donors competing with H2O2 for cpd I. As a result of the excess of TMB, the enzyme is rapidly converted to the native form, because only reversibly inactive HRP forms were produced, thus showing no decrease of activity after short incubation times of HRP with H2O2, although the heme group was already significantly affected. The presence of Gafquat, but not dextrans, shifted the oxidation of the heme group to higher H2O2 concentrations, in particular, when Gafquat was preincubated with H2O2. Prolonged incubation of HRP with H2O2 established an equilibrium between different HRP forms, which was visible in a strong absorption band at 401-405 nm (∼85% of the intensity of the original Soret band at 401.6 nm). An excess of at least 100-fold H2O2 led to the increased formation of the irreversibly inactive form P670 and a significant decrease in the main absorption band. Because of the low extinction coefficient of P670, the intensity of 3044 Analytical Chemistry, Vol. 74, No. 13, July 1, 2002

this band generally cannot be taken as an indicator of the HRP activity.12 Thus, inactivation of HRP can be found only when the activity itself is determined, and even in the presence of only a medium excess of H2O2, a decrease to ∼50% was observed. Thus, after 24 h, even an excess of electron donor (TMB) could not regenerate the inactive HRP forms to native HRP. The presence of Gafquat 755N protected HRP only slightly from inactivation. Modeling of the interaction of HRP with subunits of the polymers DEAE-dextran and Gafquat revealed no difference between the polymers, because both polymers should cover the protein surface almost completely. Monitoring the reactions of H2O2 with the enzyme via UV spectra, however, showed significant differences, in particular, with respect to the short-term protection of the heme group against H2O2 attack. Moreover, it was observed that a preincubation of only Gafquat 755N with H2O2 for 24 h reduced the H2O2 effects, depending on the concentrations of Gafquat and H2O2. That is why the coverage of the protein surface by polymers cannot be used to explain these effects. However, electrochemical determination of H2O2, that is, oxidation of H2O2 at 600 mV using a Pt electrode, showed no decrease of H2O2 concentration after 24 h incubation with Gafquat (1 mM H2O2 and 0.66% Gafquat, data not shown). Additionally, mass spectrometric analysis of mixtures of H2O2 with N-ethyl-2-pyrrolidinone or N-vinyl-2-pyrrolidinone, which resemble one of the subunits of Gafquat, showed no chemical reaction products (incubation for 24 h); however, a significant stabilizing effect combined with a short-term protection of the Soret band was observed for ami-

nopyrine (Figure 1c) and also for aminoantipyrine (data not shown), both of which have some similarities to the pyrrolidone subunit of Gafquat 755N. Thus, the stabilizing behavior of polymers can be achieved by different types of side chains, one responsible for the wrapping of the protein structure and thus stabilizing protein solutions, the other involved in a more chemical protection against oxidative or radical attack. The latter aspect could probably be improved by side chains more resembling the structure of HRP substrates, such as aminopyrine or aminoantipyrine. Thus, additional experiments with modified polymers are to be performed to improve the protection of HRP from H2O2 attack by polymers and for a more

detailed understanding of the different effects of the presently used polymers. ACKNOWLEDGMENT The authors acknowledge financial support given through the DIAMONDS project by the biotechnology program of the European Commission (no. BI04-CT97-2199). The technical assistance of S. Weissflog is also gratefully acknowledged. Received for review July 20, 2001. Accepted December 21, 2001. AC0108111


3045

Mechanistic and Molecular Investigations on Stabilization of

Recommend Documents