Biomacromolecules 2001, 2, 456-462
456
Combinatorial Screening for Enzyme-Mediated Coupling. Tyrosinase-Catalyzed Coupling To Create Protein-Chitosan Conjugates Tianhong Chen,†,‡ Rafael Vazquez-Duhalt,§ Chi-Fang Wu,| William E. Bentley,†,| and Gregory F. Payne*.†,‡ Center for Agricultural Biotechnology, 5115 Plant Sciences Building, University of Maryland Biotechnology Institute, College Park, Maryland 20742; Department of Chemical and Biochemical Engineering, University of Maryland, Baltimore County, 1000 Hilltop Circle, Baltimore, Maryland 21250; Instituto de Biotecnologia UNAM, Apartado Postal 510-3, Cuernavaca, Morelos, 62250 Mexico; and Department of Chemical Engineering, University of Maryland, College Park, Maryland 20742 Received November 20, 2000; Revised Manuscript Received March 15, 2001
In nature, tyrosinase-generated o-quinones are commonly involved in processes that lead to functional biomaterials. These biomaterials are chemically complex and have been difficult to analyze. Furthermore, the cascade of reactions involving o-quinones is poorly understood, and it has been difficult to mimic ex vivo for materials processing. We report the use of a combinatorial approach to learn how tyrosinase and low molecular weight phenolic precursors can be used to generate biologically active protein-polysaccharide conjugates. Specifically, we screened various phenolic coupling precursors and various reaction conditions for the coupling of proteins onto the polysaccharide chitosan. Several natural phenols were identified as appropriate precursors for the coupling of polyhistidine tagged organophosphorus hydrolase (His-OPH) onto chitosan films. OPH activity was retained upon coupling and subsequent studies indicated that the histidine tag was not necessary for coupling. Using conditions identified for His-OPH coupling, we observed that various biologically active proteins (cytochrome c, OPH, and His-CAT) could be coupled onto chitosan films. The glycosylated protein horseradish peroxidase was not effectively coupled onto chitosan under the conditions studied. In all cases studied, we observed that coupling required a phenolic precursor, suggesting that tyrosinase is unable to couple by reaction with surface tyrosyl residues of the target protein. In conclusion, this study illustrates a combinatorial approach for the “discovery” of conditions to couple biologically active proteins onto chitosan through natural, quinone-based processes. Introduction There is considerable interest in exploiting natural polymer processing operations to create high performance and environmentally friendly polymers. Proteins are among the best-studied biopolymers, and biochemical steps for polypeptide synthesis are well characterized and can be readily controlled. In many cases, however, proteins also undergo posttranslational processing. For instance, eukaryotic proteins often undergo glycosylation reactions that are less wellcharacterized and more difficult to control than polypeptide synthesis. Additional posttranslational processes that have largely eluded characterization and control are processes that lead to protein cross-linking or the coupling of proteins to biomaterials. These processes often exploit a single enzyme to generate a freely diffusible reactive species that undergoes subsequent nonenzymatic coupling reactions. In our work, we studied a biochemical process that is initiated by the * Corresponding author. Telephone: (301)-405-8389. FAX: (301)-3149075. E-mail:
[email protected]. † University of Maryland Biotechnology Institute. ‡ University of Maryland, Baltimore County. § Instituto de Biotecnologia UNAM. | University of Maryland, College Park.
tyrosinase-mediated conversion of phenols into reactive o-quinones. Below, we highlight two examples in which tyrosinases function in posttranslational protein processing to create functional biomaterials. Quinone reactions are integral to the ability of marine animals (e.g., mussels) to adhere to wet and submerged surfaces.1,2 These animals secrete a polyphenolic adhesive protein that consists of conserved decapeptide repeats that are rich in lysine, hydroxyproline, and dihydroxyphenylalanine (DOPA) residues.3 Some of the DOPA residues appear to be important for the attachment of the protein to the surface to confer adhesive strength.4,5 Other DOPA residues are enzymatically oxidized into o-quinones by secreted tyrosinases. These quinone residues undergo nonenzymatic cross-linking reactions2,6 that result in the formation of a protein gel7,8 that is believed to confer cohesive strength to the proteinaceous adhesive.9 Originally, it was believed that the o-quinone residue undergoes cross-linking reactions with the -amino group of lysine residues. However, chemical characterization studies have been unable to detect such linkages10 while recent studies with synthetic proteins indicate that lysine residues are not required for gel formation.4 In fact, synthetic DOPA-glutamate copolymers were
10.1021/bm000125w CCC: $20.00 © 2001 American Chemical Society Published on Web 04/27/2001
Protein-Chitosan Conjugates Scheme 1
oxidatively cross-linked as effectively as DOPA-lysine copolymers.4 A second example of quinone-based biopolymer processing in nature is cuticle sclerotization in insects (i.e., the hardening of the insect shell). During sclerotization, the enzyme tyrosinase is believed to oxidize low molecular weight sclerotizing precursors11-14 into o-quinones and these quinones undergo nonenzymatic “quinone tanning” reactions that lead to a hardened outer integument. Currently, it is believed that histidines are the most important protein residues involved in quinone tanning.15-18 Presumably, histidines are more reactive than the -amino group of lysines because histidine has a lower pKa (6.0 for histidine vs 10.5 for lysine). The above examples illustrate that quinone-based processes are important in conferring functionality to biomaterials. However, these processes remain poorly characterized because a “quinone tanned” biomaterial is a complex matrix that is difficult to analyze. Further, quinones can undergo a variety of reactions and even studies with simple model compounds have reported a diversity of products.19,20 The objective of our study was to learn how tyrosinasegenerated quinones can be used to couple a biologically active protein onto the amine-containing polysaccharide, chitosan. As illustrated in Scheme 1, our approach is based on insect sclerotization in that we use tyrosinase to oxidize a low molecular weight phenolic precursor into a reactive o-quinone that undergoes subsequent nonenzymatic reactions to couple the protein onto chitosan. Chitosan was chosen because it has amino groups with moderately low pKa’s (6.0-6.5),21,22 and because chemically coupled proteinchitosan conjugates offer interesting functional properties.23 To overcome the uncertainties in quinone coupling chemistries, we used a “discovery-based” combinatorial approach to rapidly screen various phenolic coupling precursors and reaction conditions.
Biomacromolecules, Vol. 2, No. 2, 2001 457
Materials and Methods Chemicals and Protein Preparations. Mushroom tyrosinase (EC 1.14.18.1), horseradish peroxidase (type X; EC 1.11.1.7), horse heart cytochrome c, pinacyanol chloride, paraoxon, crabshell chitosan, and acetyl CoA were purchased from Sigma Chemicals (St. Louis, MO). The phenolic coupling precursors, caffeic acid, chlorogenic acid, arbutin, dopamine, and gallate esters were obtained from Aldrich Chemical Co. (Milwaukee, WI). Buffer salts and hydrogen peroxide were purchased from J. T. Baker (Phillipsburg, NJ). The organophosphorus hydrolase containing a histidine tag (His-OPH) was obtained from Escherichia coli BL21 carrying plasmid pOPH.24 The his-tagged OPH was purified by immobilized metal affinity chromatography with TALON resins (Clontech). To study the coupling of OPH without a histidine tag, the polyhistidine tag of His-OPH was removed using enterokinase cleavage capture kit (Novagen). Recombinant chloramphenicol acetyltransferase containing a histidine tag (His-CAT) was expressed in E. coli (strain BL21pTrcHisCAT)24 and purified on an immobilized metal affinity column (IMAC Amersham Pharmacia Biotech) following the manufacturer’s procedure. Enzyme-Mediated Coupling. Proteins (His-OPH, OPH, horseradish peroxidase, cytochrome c, and His-CAT) were coupled onto a chitosan film by tyrosinase-mediated reaction in the presence of phenolic coupling precursor as illustrated in Scheme 2. Chitosan films were formed at the bottom of each well (2 cm2) of a 24-well polystyrene plate (Ultralow cluster Costar 3473) by the following procedure. Chitosan flakes (2 % w/w) were dissolved in aqueous solution by intermittently adding aliquots of 2 M HCl to maintain a pH between 2 and 3. After several hours of mixing, these solutions were filtered to remove undissolved particles, and 0.5 mL aliquots of these chitosan solutions were added to each well of the 24 well plates. The plates were then dried overnight at 60 °C to coat the bottom of the wells with the chitosan film. After being dried, the films were neutralized by contacting them with 1 mL of 1 M NaOH for about 1 h. Neutral chitosan films are insoluble, and these insoluble films were extensively washed with water and 100 mM phosphate buffer. The pH of this wash buffer was chosen to match the pH of the subsequent experiment. For protein coupling, the following additions were made to each well to bring the total volume to 1 mL; phosphate buffer (100 mM), a solution containing the target protein to be coupled, a solution containing the phenolic coupling precursor, and a solution containing tyrosinase. The concentrations and protein activities for individual experiments are listed in the legends of each Figure and Table. The 24-well plates were then incubated overnight at 4 °C on an orbital shaker. It should be noted that the tyrosinase-catalyzed coupling reaction leads to substantial “browning” of both the solution and the chitosan film, and thus standard colorometric methods for analysis could not be applied after coupling. Specifically, it was not possible to assay the protein concentrations on the films or in the supernatants, and it was not possible to measure the uncoupled activities remaining in the supernatant.
458
Biomacromolecules, Vol. 2, No. 2, 2001
Chen et al.
Scheme 2
Table 1. Combinatorial Screen of Phenolic Coupling Precursorsa
After the coupling reaction, the films were washed extensively to remove physically bound, colored material. In early experiments, washing was performed with 100 mM phosphate buffer (pH 8). In later experiments, washing was performed first with pH 9 buffer and then with pH 8 buffer in order to more completely remove physically bound colored materials (presumably oligomeric phenols) that interfered with, and reduced the sensitivity of subsequent bioassays. Tyrosinase Activity. Tyrosinase was reported by the manufacturer to have specific activities of 10 000 U/mg. We periodically measured tyrosinase activity and adjusted our additions to achieve the total activities listed in the legends to each Figure and Table. The reaction mixture for the tyrosinase assay contained tyrosinase and 0.3 mM L-tyrosine in phosphate buffer (100 mM; pH 6.5),25 and pure oxygen was bubbled through the solution. The reaction was monitored spectrophotometrically at 280 nm ( ) 3000 M-1 cm-1). After an initial lag phase, the absorbance was observed to increase linearly with time and the activity was estimated from the slope in this linear region. Protein Activities. The catalytic activities of the target proteins were estimated before coupling and the activities of the chitosan films were measured after coupling. The OPH activity was calculated as the p-nitrophenol produced from a 1 mL solution of 200 µM paraoxon in a 60 mM phosphate buffer (pH 8.5) containing 100 mM NaCl. The p-nitrophenol production was monitored spectrophotometrically at 400 nm ( ) 17 000 M-1 cm-1).26 The biocatalytic activity of cytochrome c was estimated by the decolorization of a pinacyanol chloride solution (5 µg/mL) in 10% acetonitrilephosphate buffer (pH 6) in the presence of 1 mM hydrogen peroxide.27 The extent of this decolorization reaction was estimated spectrophotometrically at 603 nm with an extinction coefficient of 82 350 M-1 cm-1. Horseradish peroxidase activity was calculated for the oxidation of a 16 mM guaiacol solution in 100 mM phosphate buffer (pH 6) in the presence of 1 mM hydrogen peroxide. Guaiacol oligomerization was monitored at 470 nm and quantified using an extinction coefficient of 26 600 M-1 cm-1.28 The CAT activity was measured spectrophotometrically by monitoring the conversion of chloramphenicol and acetyl-CoA as described by Rodriguez and Tait.29 Unless otherwise noted, all experiments were performed in triplicate.
methyl gallate chlorogenic acid gallic acid arbutin caffeic acid p-benzoquinoned catechol p-cresol propyl gallate p-methoxyphenol 3-(4-hydroxyphenyl)propionic acid dopamine 4-hydroxycinnamic acid guaiacol phenol tert-butylcatechol
OPH activity on film × 103 (U/cm2)b
coupling efficiencyc (%)
2.26 ((0.08) 2.12 ((0.14) 0.58 ((0.01) 0.54 ((0.10) 0.52 ((0.06) 0.46 ((0.03) 0.16 ((0.02) 0.12 ((0.01) 0.10 ((0.07) 0.08 0.07 ((0.05) 0.06 ((0.00) 0.05 ((0.02) 0.04 ((0.01) 0.02 ((0.03) 0.02 ((0.01)
37.8 35.4 9.7 9.1 8.6 7.7 2.7 2.0 1.7 1.3 1.1 1.1 0.9 0.6 0.4 0.3
a Coupling was studied in 1 mL buffer (pH 8) using a phenolic precursor (30 mM), tyrosinase (6 U), and OPH (0.012 U). b Standard deviation for three independent replicates except for p-methoxyphenol. c Coupling efficiency is the percentage of initial OPH activity coupled onto chitosan. d No tyrosinase was added for coupling with p-benzoquinone.
Results and Discussion Scheme 2 illustrates the combinatorial approach used in our study. Chitosan films were first deposited at the bottom of 24-well plates and coupling was achieved by incubating these films with the target protein (typically His-OPH), the phenolic coupling precursor, and tyrosinase. After the coupling reaction, the films were washed extensively to remove physically bound, colored material. After washing, each well was then assayed for activity of the target protein. Results are normalized in terms of biocatalytic activity per surface area of the chitosan film (e.g., units of OPH activity per cm2) and as a percentage of initial activity that had been coupled (see Materials and Methods for further details). It should be noted that this combinatorial approach is aimed to “discover” conditions that lead to high (or low) coupling while the reasons for high (or low) coupling may not be apparent. In the first phase of our study, we screened various phenols (30 mM) for their ability to couple His-OPH to the chitosan film. Table 1 shows that the phenols can be categorized into three groups based on their coupling abilities. The first group of phenols coupled less than 0.12 × 10-3 U/cm2 of OPH activity onto the chitosan films. We can speculate that poor
Protein-Chitosan Conjugates
coupling could result for three reasons. First, a phenol may be a poor coupling precursor if it is a poor substrate for tyrosinase (e.g., guaiacol and p-hydroxycinnamic acid). Second, a phenol may be a poor coupling precursor if the tyrosinase-generated quinone lacks two reactive sites that are accessible for coupling (e.g., the bulky tert-butyl group may limit tert-butyl-catehol’s ability to serve as a coupling precursor). Third, a phenol may be a poor coupling precursor if the tyrosinase-generated quinone is too reactive such that intramolecular reactions or oligomerization reactions compete with coupling (e.g., dopamine).30,31 Table 1 shows that p-cresol and catechol are intermediate in their coupling activities and reactions with these phenols led to (0.12-0.16) × 10-3 U/cm2 of His-OPH coupling. This observation was surprising because previous results suggested that the p-cresol-derived quinones could serve as chitosan cross-linking agents,32 and thus we anticipated that p-cresol would be an appropriate coupling precursor. The final group of phenols in Table 1 led to the coupling of over 0.5 × 10-3 U/cm2 of His-OPH to chitosan. The best two coupling agents in Table 1 are chlorogenic acid and methyl gallate. Previously, we observed that chlorogenic acid can be enzymatically grafted onto chitosan using tyrosinase, although only a low degree-of-substitution was achieved.33 The ability of methyl gallate to serve as a good coupling precursor is surprising since gallate esters have been reported to have inhibitory activities toward tyrosinase.34 The most important conclusions from Table 1 are that tyrosinase can be used to couple His-OPH to chitosan and that the coupled His-OPH retains biological activity. The controls from the experiments (not shown in Table 1) indicate that effective His-OPH coupling requires both tyrosinase and a low molecular weight phenolic coupling precursor. There is no evidence that tyrosinase alone (without coupling precursor) is capable of oxidizing surface tyrosine residues of His-OPH. Thus, coupling appears to require the generation of a low molecular weight quinone species. Table 1 also shows that phenols have varying abilities to couple His-OPH to chitosan. Interestingly, the five most effective coupling precursors were all natural products, or simple derivatives of natural phenols (i.e., methyl gallate). Finally, it should be noted that it would have been impossible to predict a priori which phenols would be the best coupling precursors. Thus, the combinatorial approach provided an efficient means to screen various phenols. To test the effect of pH on His-OPH coupling, we varied the pH of the coupling step. That is, methyl gallate, tyrosinase, and His-OPH were incubated in chitosan-containing wells at differing pH. Figure 1a shows that the His-OPH coupling to chitosan was optimal under neutral conditions. The effect of pH on coupling is expected to be complex. First, the coupling pH must be selected to account for the pH-dependencies of the activities and stabilities of the coupling enzyme (tyrosinase) and the target protein (HisOPH in this case). Figure 1b shows that His-OPH is unstable at low pH, and that its half-life increases with increasing pH. Figure 1c shows that tyrosinase activity is low under acidic conditions and is most active at slightly alkaline pH. It should be noted that tyrosinase activity was measured using
Biomacromolecules, Vol. 2, No. 2, 2001 459
Figure 1. Effect of pH on (a) His-OPH coupling, (b) His-OPH stability, and (c) tyrosinase activity. Coupling was performed using methyl gallate (30 mM), tyrosinase (6 U), and His-OPH (0.014 U). The stability of His-OPH is reported as the half-life for the first-order decay of OPH activity for the soluble enzyme. Tyrosinase activity was measured using standard conditions with tyrosine as the substrate.
standard conditions with tyrosine as the substrate25 and that the pH-activity profile may vary with different phenolic substrates. The pH can also affect coupling because the reactivities of the coupling sites on the target protein are expected to be very sensitive to pH. For instance, if Michael’s-type reactions are involved in coupling, then histidine residues can participate,17,18 and these residues would be reactive under neutral and even moderately acidic conditions because of histidine’s low pKa (6.0). If coupling requires Schiff-base formation between the quinone and the protein, then lysine residues are expected to be important, and their reactivities would increase as the pH increases toward the pKa of the -amino group (pKa ) 10.5). Alternatively, it is possible that quinone coupling involves reactions between the quinone and surface tyrosine residues of His-OPH. Because the possible coupling reactions with surface tyrosines are unclear, it is difficult to speculate on their possible pH dependencies. To examine the effect of methyl gallate concentration on His-OPH coupling, we performed a second round of combinatorial screening. Figure 2 shows that at low methyl gallate concentrations, His-OPH coupling increased with methyl gallate. Presumably this increase in coupling reflects the need for adequate levels of the coupling precursor. Above 20 mM methyl gallate, however, Figure 2 shows that
460
Biomacromolecules, Vol. 2, No. 2, 2001
Figure 2. Effect of methyl gallate concentration on His-OPH coupling. Coupling was performed in phosphate buffer (100 mM; pH 7.5) using tyrosinase (6 U), and His-OPH (0.0082 U) and varying levels of methyl gallate. The maximum activity of coupled enzyme (0.86 × 10-3 U/cm2) corresponds to the coupling of 21% of the initial OPH activity added to the reaction mixture.
Figure 3. Effect of tyrosinase level on His-OPH coupling. Coupling was performed in phosphate buffer (100 mM; pH 7.5) using methyl gallate (20 mM), His-OPH (0.012 U), and varying levels of tyrosinase. The highest activity of coupled enzyme (2.7 × 10-3 U/cm2) corresponds to the coupling of 45% of the initial OPH activity added to the reaction mixture.
coupling decreased with added methyl gallate. The reason for this decreased coupling is not clear but may result if methyl gallate is a substrate inhibitor of tyrosinase34 or if excess methyl gallate undergoes unproductive oligomerization reactions. Alternatively, the decrease in coupling at high methyl gallate concentrations could result if excess methyl gallate is converted to excess quinones that react to inactivate His-OPH. To examine the effect of tyrosinase and His-OPH levels on coupling, we performed additional rounds of screening. As shown in Figure 3, His-OPH coupling increases monotonically with tyrosinase level. Figure 4 shows that as more His-OPH is added to the reaction mixture, a larger amount of OPH activity is coupled to the chitosan film although the coupling efficiency decreases with increased initial His-OPH. Presumably, the coupling efficiency could also be altered by adjusting the surface area of the chitosan film (i.e., a greater coupling efficiency would be expected if more chitosan surface area were available).
Chen et al.
Figure 4. Effect of His-OPH level on coupling. Coupling was performed in phosphate buffer (100 mM; pH 7.5) using methyl gallate (20 mM), tyrosinase (22.5 U), and varying levels of His-OPH. The coupling efficiency is the ratio of OPH activity coupled to chitosan relative to the initial OPH activity added to the reaction mixture. Table 2. Coupling of Five Different Proteins onto Chitosana
proteins
protein activity on film × 10 3 (U/cm2)
coupling efficiency (%)
His-OPH (full reaction) without tyrosinase and phenol without tyrosinase without phenol OPH (full reaction) without tyrosinase and phenol cytochrome c (full reaction) without tyrosinase and phenol without tyrosinase without phenol His-CAT (full reaction) without tyrosinase and phenol horseradish peroxidase (full reaction) without tyrosinase and phenol
5.0 (( 0.4)
