Anal. Chem. 2005, 77, 6113-6117
Tweezing-Adsorptive Bubble Separation. Analytical Method for the Selective and High Enrichment of Metalloenzymes Birte M. Gerken,† Carsten Wattenbach,† Diana Linke,‡ Holger Zorn,‡ Ralf G. Berger,‡ and Harun Parlar*,†
Department of Chemical-Technical Analysis and Chemical Food Technology, Research Center Weihenstephan for Brewing and Food Quality, Technical University of Munich, Weihenstephaner Steig 23, D-85354 Freising-Weihenstephan, Germany, and Institute of Food Chemistry, University of Hannover, Wunstorfer Strasse 14, D-30453 Hannover, Germany
A novelly developed tweezing-adsorptive bubble separation (ABS) method for the enrichment of metalloenzymes (laccase C and horseradish peroxidase) is introduced. The method is based on the chelation of the enzymes’ active center and can also be applied for analysis. N-(2-Acetamido)iminodiacetic acid served as a chelator and was synthesized with an octyl unit to become ADA-C8. Laccase was enriched 13.3-fold (66.31% recovery) and HPOX 17.8-fold (85.34%) without a significant loss of enzymatic activity. To prove that the entire enzyme is tweezed at the active center, ABS trials were done using ADA-C8 already complexed with Cu2+ and Fe3+. As only marginal enrichment occurred (ER laccase, 0.17; ER HPOX, 0.44), no chelating effect was concluded. It was determined how the chelation toward the active center was directed by applying other chelators such as EDTA, NTA, N,N-dimethylaminoglycine, oxalic acid, malonic acid, adipinic acid, and tripropylamine, which are similar in structure to ADAC8. The results concluded that the chelation is 3-fold coordinated on the type 1 copper center of laccase, whereas that of HPOX only 1-fold at Fe3+ and additionally at the cationic amino acid arginine, which is also located at the active center. Tweezing-ABS has been proven to selectively and effectively enrich metalloenzymes.
The production of enzymes for technological applications usually implies their separation from the biological source. Frequently, applied are salting-out, dialysis, or ultrafiltration, which are often followed by chromatographic purification processes. Each of these steps, however, is bound to losses of enzymatic activity.1-3 To overcome this problem, an alternative method was investigated, the so-called adsorptive bubble separation (ABS).4 ABS can well be applied for trace analysis and the elimination of * Corresponding author. E-mail:
[email protected]. Phone: +49 (0)8161 71-3283. Fax: +49 (0)8161 71-4418. † Technical University of Munich. ‡ University of Hannover. (1) De Souza, C. G. M.; Peralta, R. M. J. Basic Microb. 2003, 43(4), 278-286. (2) Shin, K. S.; Lee, Y. J. Arch. Biochem. Biophys. 2000, 384, 109-115. (3) Regalado, C.; Asenjo, J. A.; Pyle, D. L. Enzyme Microb. Technol. 1996, 18, 332-339. (4) Lemlich, R. Ind. Eng. Chem. 1968, 60, 16-29. 10.1021/ac050977s CCC: $30.25 Published on Web 08/27/2005
© 2005 American Chemical Society
undesired byproducts at common analysis.5 In this respect, a variety of applications have been reported for mineral ores and hazardous metal ions, proteins, and surfactants.6-10 ABS has become important for the removal of trace metals such as cadmium, chromium, and copper.11 In principle, soluble and surface-active substances separate from aqueous solutions at a gas-liquid interface layer, which is generated by the inflow of gas.12,13 When gas (mostly air or nitrogen) is led through the liquid placed in a column via a porous glass frit, foam molds by starting first with spheric (lower column) and then polyhedral bubbles (upper column).14 A polyhedral foam is necessary for the purpose of concentrating substances. During transition, surface-active molecules concentrate at the interface gaseous bubble-liquid either due to drainage of the laminar liquid or because of collapsing lamellas, leading as well to a reflux. The foam thereafter flows into a beaker, where it disintegrates back to liquid, the so-called “foamate”. Common varied process parameters are as follows: gas flow rate, initial substance concentration, addition of surface-active substances, start volume of the matrix, column geometry, pH value, and foaming time. Surface-active substances such as carnosic acid, flavokavins, and solanidine alkaloids from respective plant materials could already be successfully enriched or eliminated.15-17 (5) Maas, K. In Adsorptive Bubble Separation Methods, Methodicum Chimicum; Korte, F., Ed.; Academic Press: New York, 1974; Vol. 1, pp 165-171. (6) Thomas, E. C. In Surfactant-Based Separation Processes, Part IV; Scamehorn, J. F., Jeffrey, H., Eds.; Marcel Dekker: New York, 1989; pp 233-258. (7) Malcolm, D. E.; Leahy, G. J.; Neville, T. M.; Stuart, K. N. Selective Ion Flotation of Gold from Alkaline Cyanide Solutions; AisIMM World Gold ‘91 Conference-Cairns, April, 21-26, 1991; pp 121-131. (8) Maruyama, H.; Suzuki, A.; Seki, H. J. Colloid Interface Sci. 2000, 224 (1), 76-83. (9) Bhattacharya, P.; Ghosal, S. K.; Sen, K. Sep. Sci. Technol. 1991, 26 (1011), 1279-1293. (10) Tharapiwattananon, N.; Scamehorn, J. F.; Osuwan, S.; Harwell, J. H.; Haller, K. J. Sep. Sci. Technol. 1996, 31 (9), 1233-1258. (11) Wilson, D. J.; Clark, A. N. In Handbook of Separation Process Technology; Rousseau, R. W., Ed.; Wiley-Interscience: New York, 1987; pp 806-825. (12) Ostwald, W.; Siehr, A. Chem.-Z. 1937, 64, 649-653. (13) Maas, K. Sep. Sci. 1974, 4, 457-465. (14) Manegold, E. Schaum, Chemie und Technik;Verlagsgesellschaft mbH: Heidelberg, 1953. (15) Backleh, M.; Ekici, P.; Leupold, G.; Coelhan, M.; Parlar, H. J. Sep. Sci. 2004, 27, 1042-1044. (16) Backleh, M.; Ekici, P.; Leupold, G.; Parlar, H. Naturwissenschaften 90/8 2003, 366-369.
Analytical Chemistry, Vol. 77, No. 19, October 1, 2005 6113
Figure 1. Two possible ways for the enrichment of metalloenzymes by tweezing-adsorptive bubble separation. The chelator seizes the metal ion in the enzyme’s active center. (A) Chelator-surfactant complex; ADA is used as a chelator, ionically bound to the surfactant CTAB. For dechelation, metal ions (M+) can be added. (B) Covalently bound ADA-C8, which enables foaming without CTAB. Dechelation is achieved by addition of protons.
Substances that show little or no surface activity can also be enriched but only when an interaction with a surfactant is established.4 For the enrichment of nonsurface-active proteins, for example, ionic bonds to a surfactant were created.18 For a more selective enrichment, proteins can be tagged with histidyl moieties to form a chelate with a surfactant via a metal ion, which is then transferable into the foam phase.19 An experimentally less elaborate creation of selective bonds can be achieved by the chelation of a metal ion in the active center of a metalloenzyme. Application of ABS in combination with a selective surface-active chelator is the subject of this work and is termed tweezing-adsorptive bubble separation (tweezing-ABS). Figure 1 shows two possible ways of how tweezing-ABS can principally be established. As to (A), the metalloenzyme is complexed with a chelator, which is ionically bound to a surfactant (X); the latter enables foaming. When the chelator is linked to the active center, the enzyme becomes inactivesto regain its activity, metal ions have to be added after the chelate is transferred into the foam to obtain dechelation. The remaining chelator and surfactant can be separated from the enzyme, for example, by ion exchange chromatography. Thus, it is preferred to apply covalently bound chelating surfactants (B), where the chelator is eliminated simply through precipitation by lowering the pH and filtration. Both approaches were considered in this work to test the efficiency of either way. EXPERIMENTAL SECTION Preparation of the Enzymatic Solution. The enzymatic solution (start solution) consisted either of laccase C (EC 1.10.3.2, from Trametes sp., obtained from ASA “Spezialenzyme”, Wolfenbuettel, Germany; 11.4 units L-1) in a fermentation broth or of demineralized water containing horseradish peroxidase (HPOX, EC 1.11.1.7, from Sigma, Taufkirchen, Germany; 40 units L-1). The fermentation broth was prepared by dissolving 30 g of glucose monohydrate, 4.5 g of L-asparagine monohydrate, 3.0 g of yeast extract, 1.5 g of KH2PO4, 0.5 g L-1 MgSO4, and 1.0 mL of mineral solution in 1 L demineralized water according to Sprecher.20 The chemicals were obtained from Fluka. (17) Backleh, M.; Leupold, G.; Parlar, H. J. Agric. Food Chem. 2003, 51 (5), 1297-1301. (18) Suzuki, A.; Yasuhara, K.; Seki, H.; Marruyama, H. J. Colloid Interface Sci. 2002, 253, 402-408. (19) Crofcheck, C.; Loisell, M.; Weekley, J.; Maiti, I.; Pattanaik, S.; Bummer, P. M.; Jay, M. Biotechnol. Prog. 2003, 19, 680-682.
6114
Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
Measurement of the Enzymatic Activity. The enzymatic activity was measured by a photometric assay at pH 4.5, 6.0, or 7.0 depending on chelation conditions using 3-ethylbenzthiazoline6-sulfonic acid as substrate. Therefore, 100 µL of the laccase solution, 800 µL of buffer, and 100 µL of substrate were mixed in a semimicrocuvette. For HPOX, 10 µL of the buffer were replaced with 10 µL of H2O2 (20 mM). Following 60 and 90 s. after the addition of the substrate, the absorbance was measured at 420 nm. Chelation and Dechelation. For the chelation of laccase and HPOX, between 5 and 100 mg mL-1 N-(2-Acetamido)iminodiacetic acid (ADA) was added to the enzymatic broth. To reach the highest possible stability of the chelate, the solution was set to pH 4.5, 6.0, or 7.0 depending on the experiment. For dechelation of the laccase-ADA chelates, the enzymatic broth was mixed with increasing concentrations of CuSO4‚5H2O until maximum enzymatic activity was recovered. For dechelation of the laccase-Noctylcarbamoylmethyliminodiacetic acid (ADA-C8) and HPOXADA-C8 chelates, the pH of the enzymatic broth was lowered to pH 3.5 by the addition of HCl. To eliminate ADA-C8, additional HCl was added until pH 3, which caused precipitation. Thus, the chelator could be eliminated from the enzymatic solution by filtration. Adsorptive Bubble Separation Experiments. For ABS, the start solutions were mixed each with trimethylammonium bromide (CTAB), CTAB-ADA, ADA-C8, or the metal-ADA-C8 chelate. The amount of 2.75 mL was placed in a cylindrical glass column (0.15- and 11-mm i.d.). Foam emerged after compressed air was introduced through a glass frit (porosity 16-40 µm) and was forced to liquefy by a continuous foam destructor through pressure difference. All trials were performed at room temperature. The enrichment (ER) and recovery (R) were obtained under optimized conditions of gas flow rate, pH of the starting solution, initial concentration of the surfactant, and foaming period. The values were calculated using the following equations:
ER )
EAf EAs
R)
EAf‚Vf ‚100 EAsVs
where EAf is the enzymatic activity in foam (units L-1), EAs the enzymatic activity in the starting solution (units L-1), Vf the volume (20) Sprecher, E. Planta 1959, 53 (6), 565-574.
Figure 2. Laccase and HPOX chelated by ADA-C8, depicted in gray with oxygen in red and amino nitrogen (N) in blue. A possible attachment of the chelator on the active center of the metalloenzymes is shown. ADA-C8 is attached to the type 1 copper (pink) of laccase with two carboxyl groups (X) and with N, whereas HPOX is bound by N on the Fe3+ sphere (yellow) and with one carboxyl group (X) on arginine (Arg); the other carboxyl group Y is uncoordinated. The depiction was performed with PyMOL software.26
of the liquefied foam (L), and Vs the volume of the starting solution (L). Chelation of ADA-C8 with External Metal Ions. ADA-C8 was reacted with external metal ions in equimolar concentrations before the addition of the enzyme. For laccase, 25 mg of CuSO4‚ 5H2O and 30 mg of ADA-C8 were dissolved in 10 mL of demineralized water under stirring. Afterward, laccase was added with an activity of 11.4 units L-1. For HPOX, 590 mg of FeCl3 was mixed with 650 mg of ADA-C8 in 10 mL of demineralized water before HPOX was added with an enzymatic activity of 40 units L-1. Solving Experiments. The enzymes were chelated with ADAC8 at pH 10 as described above (11.4 units L-1 laccase + 3.0 mg L-1 ADA-C8; 40 units L-1 HPOX + 65.0 mg mL-1 ADA-C8). The pH was lowered by the addition of HCl; the pH was measured when the first precipitation was visible. Synthesis of N-Octylcarbamoylmethyliminodiacetic Acid. ADA-C8 was synthesized by solving iminodiacetic acid (1.33 g, 10 mmol) and N-octyl-2-chloroacetamide (2.06 g, 10 mmol) in 50 mL of ethanol. Sodium hydroxide (10%) was used to adjust the solution to pH 11. The mixture was refluxed for 5 h and kept constantly at pH 11 by adding sodium hydroxide (10%). Thereafter, the solution was evaporated to dryness. The residue was diluted with 80 mL of distilled water and extracted with diethyl ether.
Concentrated hydrochloric acid (37%) was added to the aqueous phase until pH 2. The obtained precipitate was filtered and dried under vacuum. A yield of 71% (2.16 g or 7.13 mmol) was obtained. The substance was characterized by 1H NMR and 13C NMR spectroscopy: (Bruker, AC 250) 1H NMR (250 MHz, D6-DMSO) δ 0.85 (t, 3J ) 6.5 Hz, 3H, CH3), 1.24 (br m, 10H, CH2), 1.39 (m, 2H, CH2), 3.08 (m, 2H, NHCH2CH2), 3.28 (s, 2H, NCH2CO), 3.43 (s, 4H, CH2COOH), 8.02 (t, 3J ) 5.8 Hz, 1H, NH). 13C NMR (62.5 MHz, D6-DMSO) δ 13.9 (CH3), 22.1 (CH2), 26.3 (CH2), 28.6 (CH2), 28.7 (CH2), 29.1 (CH2), 31.2 (CH2), 38.2 (CH2-N), 55.4 (2C, CH2COOH), 58.0 (N-CH2COOH), 170.4 (CONH), 172.7 (2C, COOH). RESULTS AND DISCUSSION The enzyme exploited first was laccase containing a trinuclear copper cluster and an additional blue copper center (type 1), which is less than 13 Å away from the three-cored cluster (Figure 2). The blue copper shows a trigonal coordination with two histidyl and one cystyl moieties as equatorial ligands.21 The trinuclear copper cluster consists of a one-cored type 2 and a two-cored type 3 copper; the type 2 copper is coordinated by two and the type 3 by six histidyls. ADA was added for the chelation of the copper ion, and the loss of enzymatic activity caused thereby served as a measure of (21) Claus, H. Micron 2004, 35, 93-96.
Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
6115
Figure 3. Enzymatic activity of laccase vs concentration of ADA, added for chelation at pH 4.5, and CuSO4‚5H2O, added for dechelation at pH 10.
chelation. With increasing concentration of ADA, the initial activity of laccase (11.4 units L-1) decreased, reaching a complete inactivation with 55 mg L-1 at pH 4.5 (Figure 3), which meant in turn that the enzyme was completely chelated. With increasing pH, smaller amounts of the chelator were required for the inactivation, 1.9 mg mL-1 at pH 6 and 0.2 mg mL-1 at pH 7, which was ascribed to the increasing stability of the chelate. To use the chelation property of the enzyme for ABS, it had to be ensured that the chelation can be reversed to regain the enzymatic activity. Dechelation of laccase was achieved through the addition of Cu2+ ions or by lowering the pH to 3.5. After chelation and dechelation were demonstrated, ABS trials started. Their efficiency is commonly expressed by the enrichment and recovery.22 No formation of foam was observed with solutions containing laccase and ADA. As full catalytic activity of laccase was recovered from foam generated in the presence of CTAB (Figure 3), this surfactant was added assuming that it would form an ionic bond with the chelate. Compared to experiments without ADA, the enrichment of the laccase-ADA-CTAB complex via tweezing-ABS did not increase significantly, only from 3.67 to 4.12, along with Rs of 100 and 71.5%, respectively. The lower recovery of the chelated fraction was attributed to the loss of enzymatic activity during chelation and dechelation, but nevertheless, the enrichment was increased in the presence of ADA. Considering the relatively low ER, it is assumed that CTAB interacts both with ADA and with the residues of the enzyme. Furthermore, interactions with other substances available in the fermentation broth cannot be excluded. To tweeze the entire enzyme more selectively, ADA was covalently coupled with an N-octyl moiety to receive ADA-C8. The addition of CTAB for ABS was not necessary anymore, because the new chelator gave ample foam. Thereby, the enrichment was increased to 13.3 (R ) 66.3%) (Figure 4). This significantly higher enrichment was assigned to the more exclusive bonding of ADA-C8 to the active center of laccase. To substantiate these findings, HPOX was used next, an enzyme showing a ferriprotoporphyrin group in the active center containing Fe3+ (Figure 2).23 The fifth coordination position of the ion is occupied by a histidyl ligand, whereas the sixth position is free.24 Additionally, HPOX contains two Ca2+ ions near the active (22) Brown, A. K.; Kaul, A.; Varley, J. Biotechnol. Biochem. Eng. 1999, 62, 278290. (23) Howes, B. D.; Feis, A.; Raimondi, L.; Indiani, C.; Smulevich, G. J. Biol. Chem. 2001, 276, 40704-40711. (24) Gajhede, M.; Schuller, D. J.; Henriksen, A.; Smith, A. T.; Poulos, T. L. Nat. Struct. Biol. 1997, 4, 1032-1038.
6116 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
Figure 4. Enrichment ratios (ER) of laccase and HPOX chelates. For the enrichment of 1, the standard adsorptive bubble separation process was applied using CTAB as a surfactant, which interacts ionically with the enzyme. The enrichment of the chelates 2-4 occurred via tweezing-ABS, conducted with CTAB also ionically bound to the enzyme-ADA chelate (2), and with a synthesized ADA-C8 chelator (ADA covalently bound to an N-octyl moiety) to generate foam (3, 4).
center. The enzyme was likewise inactivated by the addition of ADA. A concentration of 40.0 mg mL-1 was necessary to fully eliminate the initial activity of 40 units L-1 at pH 6. A 17.7-fold enrichment and 85.4% recovery of the HPOX-ADA-C8 chelate was achieved by tweezing-ABS (Figure 4), which was even higher than for laccase. It is assumed that the high enrichment and loss of activity of both enzymes by applying tweezing-ABS was due to chelation of the metal ions in the active center by ADA or ADA-C8. To exclude that the loss of enzymatic activity is caused by an extraction of the metal ions from the active center applying ADA or ADA-C8, the latter was chelated with external Cu2+ or Fe3+ metal ions. These chelates were then subjected to the enzymatic broth and afterward applied for ABS. This resulted in only marginal enrichment and recovery rates of the enzymes: 0.14 ER and 5.8% R for laccase, and 0.44 ER and 10.84% R for HPOX. Therefore, the enzyme is not enriched by ABS without the creation of a bond between the chelator and the enzyme’s metal ion. Furthermore, there are no interactions between the Cu2+/Fe3+-ADA-C8 chelates and the enzymes. Consequently, no extraction of the enzyme’s metal ion occurs by tweezing-ABS, because if an extraction took place, the chelator would only enrich the enzyme’s metal ion in the foam but not the apoenzyme. Thus, only little enrichment of the enzyme would be possible by tweezing-ABS. This assumption was further confirmed by pH precipitation experiments using chelate solutions of Cu2+-ADA-C8, Fe3+-ADAC8, and laccase/HPOX-ADA-C8. It was found that when the pH was lowered, the metal chelates Cu2+-ADA-C8 and Fe3+-ADAC8 precipitated below pH 8, whereas the enzyme-ADA-C8 chelates remained dissolved until pH 3. Some mechanistic insights of the attachment and orientation of the chelators ADA and ADA-C8 toward the metalloenzymes’ active centers were derived from inhibition studies using different chelators (Table 1). Compared to ADA, these chelators differed in size and numbers of carboxylic groups and nitrogen atoms. While chelators such as ethylenediaminotetraacetate (EDTA) and nitrilotriacetic acid (NTA), having at least one amine and two carboxyl groups in common with ADA, inactivated both enzymes;
Table 1. Loss of Activity of Metalloenzymes due to Chelation with Different Chelators (pH 6.0; EA,laccase: 11.4 Units L-1, EA,Hpox 40 Units L-1) loss of activity (%) chelator
laccase Ca
HPOXb
ADA EDTA NTA N,N-dimethylaminoglycine oxalic acid malonic acid adipinic acid tripropylamine
100 100 100 37.5 33.3 20.0 0 ndc
100 100 100 100 60.0 55.0 0 0
a Concentration, chelator: 0.01 mmol mL-1. b Concentration, chelator: 0.21 mmol mL-1. c Not determined.
other chelators reacted specifically toward the enzyme. Thus, different ways of attachment can be assumed. For laccase, the type 1 copper is known for substrate binding and conversion, whereas the types 2 and 3 copper centers are inaccessible (Figure 2), because they can only be reached via water channels presenting a steric hindrance.25 Thus, the chelation can be focused on the central type 1 copper. It is assumed that the type 1 copper is coordinated 6-fold when ADA-C8 is applieds three coordination places are fixed through amino side chains of the enzyme, and the other three via the chelator itself. This assumption could be verified by the fact that a complete chelation was only possible with other at least three times coordinating ligands such as EDTA and NTA (Table 1). The application of two times coordinating dicarboxylic acids such as oxalic acid, malonic acid, and adipic acid resulted instead in insufficient chelation and necessitated rather higher amounts of chelators for a successful and complete inactivation of the enzyme. The acceptability of the chelating acids decreases with increasing chain length between the carboxylic groups from oxalic acid to adipic acid, which is ascribable to decreased acid strenght and steric hindrance. Another two times coordinating chelator used was N,N-dimethylaminoglycine, having, other than the dicarboxylic acids, only one amino and one carboxyl group, with which it was also not possible to completely chelate the laccase. Using N,N-dimethylaminoglycine for the chelation of HPOX, however, resulted in complete inactivation of the enzyme, which assumes in turn another way of chelation as compared to laccase. Here, the Fe3+ central atom of the porphyrin ring is already 5-fold coordinated, thus making available only one coordination place for the chelator. It is assumed that this free coordination place is taken by the nitrogen of the chelator (e.g., ADA) as it shows the (25) Enguita, F. J.; Martins, L. O.; Henriques, A. O.; Carrondo, M. A. J. Biol. Chem. 2003, 278, 19416-19425. (26) DeLano, W. L. The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos, CA, 2002.
highest affinity toward Fe3+. Furthermore, the addition of the tertiary amine tripopylamine has not chelated the enzyme, and thus, it is assumed that further interactions are built up by carboxyl groups toward the cationic amino acid arginine as part of the substrate bag and actively participating in the substrate turnover. Due to the fact that a complete chelation of the enzyme was possible with N,N-dimethylaminoglycine (having one amino and one carboxyl group) as well as with ADA, EDTA, and NTA, it is assumed that the second carboxyl group of ADA does not significantly take part at the coordination. However, chelators without nitrogen but with two carboxyl groups were also capable of chelating HPOX though not entirely. This insufficient chelation is ascribed to a weaker linkage between the carboxyl group and Fe3+ central atom compared to nitrogen. Although HPOX could be better chelated with dicarboxylic acids than laccase, the usefulness of dicarboxylic acids to inactivate the enzyme decreased with longer chain lengths between the carboxyl groups, which is also deduced from steric hindrance and decreased acid strength. As HPOX also contains two Ca2+ metallic cations, a chelation thereon seems possible, but would not necessarily lead to a loss of activity, because Ca2+ is not directly participating in the substrate conversion and is just stabilizing the active center due to its location above and below. Another aspect speaking against the chelation at Ca2+ is the coverage and accessibility of the cations. CONCLUSION Tweezing-adsorptive bubble separation, based on the selective chelation of a metal cation in the active center of an metalloenzyme, increased the foam enrichment of it. Substituting the previously used common surfactant CTAB by the novel tweezer molecule resulted in a more than 3-fold improvement of laccase foaming. The findings were repeated for HPOX with even higher enrichment. The hybrid molecule ADA-C8 proved to be more efficient than the mere mixture of ADA with CTAB (Figure 1). ACKNOWLEDGMENT Financial support by the Federal Ministry of Economics and Labor (BMWA, Germany) via the Research Association of the German Food Industry (FEI) is gratefully acknowledged (project 121 ZN). SUPPORTING INFORMATION AVAILABLE Diagram of the adsorptive bubble separation apparatus, the varied process parameters, and graphical material for more information about the chelation of the metalloenzymes. This material is available free of charge via the Internet at http:// pubs.acs.org. Received for review June 3, 2005. Accepted August 11, 3005. AC050977S
Analytical Chemistry, Vol. 77, No. 19, October 1, 2005
6117
