518
Biomacromolecules 2008, 9, 518–522
Humicola insolens Cutinase-Catalyzed Lactone Ring-Opening Polymerizations: Kinetic and Mechanistic Studies Mo Hunsen,† Azim Abul,‡ Wenchun Xie,‡ and Richard Gross*,‡ Department of Chemistry, Kenyon College, Gambier, Ohio 43022, NSF-I/UCRC Center for Biocatalysis and Bioprocessing of Macromolecules, Polytechnic University, Department of Chemistry and Chemical Engineering, Six Metrotech Center, Brooklyn, New York 11201 Received November 16, 2007
This paper explores reaction kinetics and mechanism for immobilized Humicola insolenscutinase (HIC), an important new biocatalyst that efficiently catalyzes non-natural polyester synthetic reactions. HIC, immobilized on Lewatit, was used as catalyst for -caprolactone (CL) and ω-pentadecalactone (PDL) ring-opening polymerizations (ROPs). Plots of percent CL conversion vs time were obtained in the temperature range from 50 to 90 °C. The kinetic plot of ln([M]0/[M]t) vs time (r2 ) 0.99) for HIC-catalyzed bulk ROP of CL was linear, indicating that chain termination did not occur and the propagation rate is first order with respect to monomer concentration. Furthermore, linearity to 90% conversion for Mn vs fractional CL conversion is consistent with a chain-end propagation mechanism. Deviation from linearity above 90% conversion indicates that a competition between ring-opening chain-end propagation and chain growth by steplike polycondensations takes place at high monomer conversion. HIC was inactive for catalysis of L-lactide and (R,S)-β-butyrolactone ROP. HIC-catalyzed ROP of -CL and PDL in toluene were successfully performed, giving high molecular weight poly(-caprolactone) and ω-poly(pentadecalactone). In addition, the relative activities of immobilized Candida antarctica lipase B (CALB) and HIC for -CL and PDL polymerizations are reported herein.
Introduction Next-generation materials will be more sophisticated, working with higher performance at a primary task such as coating while having secondary functions such as antimicrobial activity, reactivity with toxins, and more.1,2 To meet this vision, new synthetic methods are needed that provide cost-effective materials with increased structural complexity and that meet strict environmental impact standards. Catalysts for synthesis of nextgeneration materials will be challenged with selective transformations and should operate under mild conditions without stringent requirements for exclusion of atmospheric substances such as water and oxygen. Enzymes are natural catalysts offering a broad spectrum of chemical reactivities. They operate under mild conditions and generally tolerate the presence of water, oxygen, and other substances that deactivate chemical catalysts. For these reasons and more, enzymes are attractive options for mining next-generation catalysts to generate improved processes and material performance. Thus far, the majority of enzymes studied for polyester synthesis have been from the lipase family, with lipase B from Candida antarctica (CALB) as the dominant enzyme.3–5 Furthermore, Novozym 435, that is, CALB immobilized on Lewatit supplied by Novozymes, has shown extraordinary reactivity for a wide range of polyester forming ring-opening (ROP) and condensation polymerizations. Previous work by our laboratory established that lipasecatalyzed lactone polymerizations occur with some elements of “control”. For porcine pancreatic lipase (PPL) catalyzed -caprolactone (-CL) polymerizations, linearity of log([M]0/ * Corresponding author. E-mail:
[email protected]. † Department of Chemistry, Kenyon College. ‡ NSF-I/UCRC Center for Biocatalysis and Bioprocessing of Macromolecules, Polytechnic University, Department of Chemistry and Chemical Engineering.
[M]t) vs time plots indicated that the number of chains remained constant during polymerizations and chain termination does not occur. Furthermore, Mn vs conversion plots were linear to high monomer conversion consistent with chain growth by ringopening from chain ends as opposed to by a step-growth mechanism6,7 reported for Novozym 435 catalyzed -CL ringopening polymerizations that Mn can be varied between ∼5000 and 21000 by changing enzyme water content from 0.4 to 2.8 wt %. Thus, by judicious preselection of enzyme water content, predictable polyester molecular weights are attained. Subsequently, Mei et al.8 extended the kinetic plot of ln([M]0/[M]t) vs time to 96% -CL conversion and Mn 11970. Inspection of Mn vs -CL conversion plots at later stages of the polymerization, for reactions at low enzyme water content, showed an upward deviation from otherwise linear Mn vs fractional -CL conversion plots. This is evidence that, at high monomer conversion, step-condensation reactions between enzymeactivated carboxyl chain ends and hydroxyl terminal groups become increasingly important. Verification of CALB-catalyzed transesterification reactions between high molecular weight chains was obtained by using Novozym 435 as catalyst and preformed P(CL) and poly(ω-pentadecalactone), P(PDL), as substrates. For bulk reactions between PCL (Mn 9200) and P(PDL) (Mn ) 4300), mixed PDL*CL/CL*PDL dyads were observed by 13C NMR within 30 min. Furthermore, by increasing the reaction time, Novozym 435 catalyzed the formation of random copolymers.9 Hult and co-workers used MALDI-TOF to characterize macrocyclic species formed during CALBcatalyzed lactone ring-opening polymerizations.10 Recently, Thurecht et al.11 reported kinetics of Novozym 435 catalyzed CL polymerizations in scCO2. Relatively poor molecular weight control was attributed to a large degree of transesterification, forming both cyclic and linear products.
10.1021/bm701269p CCC: $40.75 2008 American Chemical Society Published on Web 01/17/2008
Humicola insolens Cutinase-Catalyzed Lactone ROP
Biomacromolecules, Vol. 9, No. 2, 2008 519
Ring size was shown to be an important factor influencing lipase-catalyzed lactone polymerizations. Palmans and coworkers12 summarized previous work that compared effects of lactone ring size for lipase and chemical catalyzed polymerizations. Of particular relevance to the current study are reports by Kobayashi and co-workers13,14 that, by using Pseudomonas fluorescens lipase (lipase PF), the rate at which lactones are converted to polyester increases with their ring size. Also, our laboratory reported Novozym 435 polymerizes PDL faster than -CL.15 Furthermore, Palmans and co-workers showed KM was relatively independent of the ring size, suggesting that CALB has similar affinities for these lactones. However, 7-heptanolactone (HL), 12-dodecalactone (DDL), and PDL showed significantly higher Vmax values relative to other lactone ring sizes studied. In the current study, research was begun to define effects of lactone ring size on HIC activity. Cutinases are extracellular fungal enzymes whose natural function is catalysis of cutin ester-bond hydrolysis. Cutin is a lipid-polyester found in the cuticle of higher plants.16–18 With molecular weights of around 20 kDa, cutinases are the smallest members of the serine R/β hydrolase superfamily.19 Cutinases have shown activity on a broad variety of esters including triglycerides.20 Synthetic activities of cutinases have also been described for the preparation of agrochemicals containing one or more chiral centers.21 The most studied cutinase is that from the fungus Fusarium solani pisi.20 Cutinases have been found to have good hydrolytic activity on synthetic fibers such as poly(ethylene terephthalate), suggesting their potential for biopolishing applications.22,23 Recently, our laboratory communicated that the cutinase from Humicola insolens (HIC) is a promising new biocatalyst for polyester synthesis.24 Immobilized HIC showed optimal activity at 70 °C and catalyzed condensation and lactone ring-opening polymerizations in bulk and using toluene as solvent, yielding high molecular weight polyesters. Herein we report a detailed kinetic and mechanistic investigation for lactone ring-opening polymerizations using immobilized HIC, a promising new catalyst for industrially important polyesterification reactions. Plots of Mn versus conversion were constructed to analyze the mechanism of chain growth. Also, plots of ln([Mo]/[Mt]) versus time were constructed and analyzed to probe the occurrence of termination reactions and the kinetics of monomer conversion. To explore substrate specificity of HIC, its activity was studied as a function of lactone ring size and steric hindrance. For this purpose, the activity of immobilized HIC for polymerizations of (R,S)-β-butyrolactone, (L)-lactide, δ-valerolactone (VL), CL, and PDL was assessed. Finally, relative activities of immobilized HIC and Novozyme 435 for CL and PDL ring-opening polymerizations were studied.
washed with phosphate buffer (pH 7.8, 0.1 M, 3 × 10 mL). Cutinase (used as received, 1.7 mg mL-1, 25 mL) was added to the resin and incubated with gentle shaking (100 rpm) at 4 °C for 48 h. Remaining supernatant was removed by centrifugation (10000 rpm, 5 min, 15 °C). The cutinase loaded resin was freeze-dried, and the dry weight determined (W). The protein contents of supernatants before (P1) and after immobilization (P2) were determined using the bicinchoninic acid method.25 Protein loading was calculated in mg protein g-1 using the relationship (P1 – P2)/W. Cutinase loading was 11 mg protein g-1 or 0.1% by wt. HIC cutinase physically immobilized on Lewatit OC VOC 1600 and Novozym 435, which consists of CALB physically immobilized on the identical resin, were dried over P2O5 under vacuum using a pump (0.1 mmHg; 24 h, rt) prior to use. General Procedure for Novozym 435 and HIC Catalyzed Polymerization. Lactone (1 g), toluene (2 mL, for reactions not run in bulk), and 100 mg of immobilized HIC (0.1% w/w cutinase to total weight of monomer) was added under nitrogen atmosphere in a parallel reactor (Argonaut Advantage Series 2050). When using Novozym 435, 100 mg of catalyst was added, giving a CALB-to-monomer ratio of 1% w/w. Reactions were stirred magnetically while maintaining temperatures at predetermined values (50-90 °C). An aliquot was withdrawn at specified time intervals, chloroform was added, catalyst beads were removed by filtration, solvent was evaporated, and 1H NMR spectra were recorded in CDCl3. NMR solvent was then evaporated and GPC was recorded on the same sample using THF as eluent for P(CL) and poly(VL) and CHCl3 for P(PDL). Reactions were terminated at 24 h by adding CHCl3 (20 mL), filtration to remove catalyst beads, stripping solvent by rotoevaporation, and drying in a vacuum oven. Products were analyzed as above by 1H NMR and GPC. Instrumentation Methods. Nuclear Magnetic Resonance (NMR). -CL polymerizations were monitored by 1H NMR to determine monomer conversion. 1H NMR spectra were recorded in CDCl3 on a Bruker NMR spectrometer (model DPX300) at 300 MHz. The chemical shifts in parts per million (ppm) for 1H NMR spectra were referenced relative to tetramethylsilane (TMS, 0.00 ppm) as the internal reference. The unique signals at 4.23 (t, OC H2) and 2.64 (t, C(O)C H2) were assigned to protons of CL monomer. Signals at 4.06 (t, OC H2), 3.64 (t, HOC H2), and 2.30 (t, C(O)C H2) appeared after the onset of polymerization reactions and were assigned to P(CL) protons. The ratio of signals at 4.06 to 4.23 was used to calculate monomer conversion and Mn, respectively. Gel Permeation Chromatography (GPC). The number- and weightaverage molecular weights (Mn and Mw, respectively) were determined by size exclusion chromatography using a Waters 510 pump, a 717 plus autosampler, and a Wyatt Optilab DSP interferometeric refractometer coupled to 500, 103, 104, and 105 Å Ultrastyragel columns in series. Trisec GPC software version 3 was used for calculations. THF (CHCl3 for P[PDL]) was used as eluent with a flow rate of 1.0 mL min-1 at 35 °C. Molecular weights were determined on the basis of a conventional calibration curve generated by narrow molecular weight polystyrene standards obtained from Aldrich Chemical Co.
Experimental Section
Results and Discussion
Polymerization grade -caprolactone (CL), a gift from Union Carbide, was first dried over calcium hydride and then distilled under reduced pressure in a nitrogen atmosphere. Toluene-d8, δ-valerolactone (VL), and ω-pentadecalactone (PDL) were purchased from Aldrich Chemical Co. Humicola insolens cutinase and Novozym 435 (immobilized Candida antartica lipase B, CALB, specified activity 10000 PLU/g) were gifts from Novozymes. All liquid chemical transfers were performed by syringe through rubber septum caps under a nitrogen atmosphere. All other solvents and reagents were obtained commercially at the highest purity available and used without further purification. Immobilization of Humicola insolens Cutinase (HIC). Humicola insolens cutinase (HIC) was immobilized on Lewatit OC VOC 1600. The resin (3 g) was first activated with ethanol (10 mL) and dried under vacuum for 60 min to remove traces of ethanol. The resin was further
Humicola insolens Cutinase Catalyzed Bulk Ring-Opening Polymerization of E-Caprolactone. Immobilization of Humicola insolens Cutinase. Humicola insolens cutinase (HIC) was physically immobilized on Lewatit OC VOC 1600, a macroporous poly(methyl methacrylate) resin (see Experimental Section). The protein loading efficiency was 78%, resulting in 11 mg of cutinase per g catalyst. Polymerizations described below were performed with 100 mg of immobilized cutinase and 1 g of monomer. Thus, the w/w ratio of cutinase-tomonomer was maintained at 0.1% by wt. Effect of Temperature on HIC-Catalyzed Bulk RingOpening Polymerization of E-CL. Previously, we reported HIC catalyzed -CL ROP at one reaction time (24 h).24 Here, time-
520 Biomacromolecules, Vol. 9, No. 2, 2008
Hunsen et al.
Figure 1. Monomer conversion vs time for HIC-catalyzed CL ringopening polymerizations in bulk at different temperatures. Figure 3. Semilogarthmic plots for HIC-catalyzed CL ring-opening polymerizations in bulk at different temperatures.
Figure 2. Plots of Mn vs %-CL conversion for CL ring-opening polymerizations in bulk at different temperatures.
course studies of bulk -CL polymerizations were performed at different temperatures in the range 50-90 °C. Figure 1 shows corresponding plots of percent CL conversion vs reaction time. The optimal temperature for HIC-catalyzed bulk ROP of CL is 70 °C. A large reduction in HIC activity for CL polymerization was found when the reaction temperature was increased from 70 to 80 °C. Further increase of reaction temperature to 90 °C resulted in complete loss of HIC activity for this reaction. This may be due to a change in HIC conformation or denaturation at temperatures greater than 70 °C that negatively impacts HIC’s catalytic activity. Monomer conversions greater than 80% within 32 h were attained only for polymerizations at 60 and 70 °C. On the basis of these results, all further studies herein on lactone polymerizations were conducted at 70 °C. Proton nuclear magnetic resonance (NMR) spectroscopy was used to determine percent monomer conversion, and gel permeation chromatography (GPC) gave values of molecular weight averages (Mn, Mw) and polydispersity index (Mw/Mn, PDI). At 70 °C, monomer conversion reached 99% by 24 h. The highest Mn for HIC-catalyzed CL bulk ROP were 28000 (Mw/Mn 2.1) at 70 °C and 48 h. Linearity of Mn versus conversion plots for HIC-catalyzed CL ring opening polymerizations at 50, 60, 70, and 80 °C (Figure 2) indicates chain growth occurs by propagation at chain ends to high conversion. This mechanism of chain growth was similarlyseenforPPL-andCALB-catalyzedCLpolymerizations.6,8 In other words, intriguingly, two lipases from different species and a cutinase all share this common chain-growth mechanism. The plot of Mn versus conversion (Figure 2), constructed by taking together results from polymerizations at 50, 60, 70, and
80 °C, fall within a single line to 90% conversion. Therefore, although temperature has a large effect on the rate of monomer conversion, temperature had no apparent effect on product molecular weights taken at the same monomer conversion values. This is explained by no changes in the rate of initiation or other mechanistic features such as termination reactions that would alter product molecular weight at similar conversion values. Therefore, similar to PPL- and CALB-catalyzed CL polymerizations, HIC-catalyzed CL ROP allows control of product molecular weight via regulation of monomer conversion.6,8 The large increase in Mn for conversions >90% at 70 °C shows that HIC-catalyzed step condensation reactions between enzymeactivated carboxyl chain ends and hydroxyl terminal groups contributes to molecular weight increase at high monomer conversions (Figure 2). A similar behavior was observed previously by our group for Novozyme 435-catalyzed ROP of CL.8 To determine whether kinetics of chain growth are first order with respect to monomer conversion, and whether events of chain termination occur, semilogarithmic plots of ln[Mo]/[Mt] versus time were constructed for HIC-catalyzed bulk ROP’s of -caprolactone (Figure 3). Except for 80 °C, the data are well described by linear regression analysis (r2 g 0.99). Therefore, at 50, 60, and 70 °C, kinetics of chain growth is first order with respect to monomer conversion and events of chain termination did not occur. As above, this mechanistic feature was similarly observed for PPL- and CALB-catalyzed CL polymerizations.6,8 The lack of termination events at 50, 60, and 70 °C is consistent with the observation above that polymer molecular weight at similar conversion is independent of temperature, at least between 50 and 70 °C. The turnover number (TON, mol of CL converted to polymer per mol of cutinase per min) for the above polymerizations was determined from the initial slope of %-monomer conversion versus time plots. Figure 4 displays the resulting plot of TON versus reaction temperature. A maximum in the plot is observed at 70 °C where the TON is 770. Studies were performed to determine whether immobilized HIC is active on lactones that differ from CL in ring-size and substitution. Nonsubstituted aliphatic lactones VL, CL, and PDL, with ring sizes of 6, 7, and 16 were selected (Scheme 1). Methyl substituted lactones investigated include β-butyrolactone (BL) and L-lactide, with ring sizes of 4 and 6, and methyl substituents β- and R- to lactone carbonyls, respectively. Polymerizations
Humicola insolens Cutinase-Catalyzed Lactone ROP
Biomacromolecules, Vol. 9, No. 2, 2008 521 Table 1. Ring-Opening Polymerization of Lactones of Differing Structure Catalyzed by Novozym 435 and Immobilized HICa monomer
catalyst
Mn
PDI
CL CL VL BL L-lactide PDL PDL
Novozym 435 HIC HIC HIC HIC Novozym 435 HIC
29400 24900 1700 ndb nd 43100 44600
1.5 1.7 1.45 nd nd 1.9 1.7
a Reactions performed for 24 h, under N2, at 70 °C; lactone-to-toluene was 1:2 w/v; monomer to protein for immobilized HIC and Novozym 435 was 0.1 and 1% w/w, respectively. b nd is not determined due to nearly quantitative recovery of monomer from reactions.
Figure 4. Plot of turnover number (TON, mol of monomer converted to product per mol of HIC per min) vs temperature for HIC-catalyzed caprolactone bulk ring-opening polymerization at different temperatures. Scheme 1. HIC-Catalyzed Lactone Ring-Opening Polymerizations
As above, polymers formed were not fractionated by precipitation prior to analysis by GPC. No peak corresponding to monomer was observed in GPC chromatograms of all products, indicating that monomer conversions were >97%. Inspection of Table 1 shows that synthesis of P(CL) and P(PDL) with immobilized HIC and Novozym 435 gave polymers of similar values of Mn and PDI despite the large difference in monomer to protein ratios. Work is underway to define differences in kinetics for these two catalyst systems. This will require preparation of immobilized HIC and CALB catalysts where protein loading is similar. Otherwise, differences in intra- and interparticle diffusion rates can be expected at different protein loadings that will ultimately influence comparisons between these two protein catalysts.
Summary and Conclusions
were conducted for 24 h, at 70 °C, in toluene, with 0.1% w/w cutinase relative to monomer. HIC-catalyzed VL and CL polymerizations gave products with Mn (Mw/Mn) of 2700 (1.45) and 24900 (1.7), respectively. Hence, it appears that increase in the ring-size from 6 to 7 resulted in a large increase in substrate reactivity for polyester synthesis. This result differs from that obtained with Novozym 435 catalyzed polymerization of VL and CL where KM was nearly identical, Vmax for CL was larger by a factor of about 2, and Mn (Mw/Mn) was 19000 (4.0) and 10000 (2.2), respectively. HIC was inactive for polymerizations of the methyl substituted lactones BL and L-lactide, resulting in nearly quantitative recovery of monomer. While no reports of successful polymerization of L- or L,D-lactide by Novozym 435 were found, the lipase from Pseudomonas cepacia (lipase PS) appears active for polymerizations of L,Land D,L-lactide.26,27 In fact, Matsumura reported that by using 3% by wt of lipase PS relative to D,L-lactide poly(D,L-lactide) was obtained that after precipitation had an Mw of 69000. With respect to enzymatic polymerization of β-butyrolactone, early studies by Nobes et al.28 using porcine pancreatic lipase and lipase PS in-bulk or low polarity organic media, as well as recent work by Gorke et al.,29 using Novozym 435 in ionic liquids gave only oligomers of β-butyrolactone. A direct comparison of polymers obtained by ring-opening polymerizations of -CL and PDL catalyzed by HIC and Novozym 435 was performed. Reaction conditions were identical (1:2 w/v substrate-to-toluene, 70 °C, 24 h) except that the % by wt of monomer to enzyme was 0.1 and 1%, respectively.
This work was motivated by the need to identify new enzymes for polymer synthesis that can provide variable specificity and other valuable attributes. For HIC, immobilized on Lewatit, its optimal activity for lactone ring-opening polymerizations is 70 °C. Chain transfer to monomer does not occur and monomer conversion followed first-order kinetics. Also, the kinetic analysis performed indicated that, to high monomer conversion, chain growth takes place by propagation at chain ends and chain termination did not occur. However, with depletion of lactone monomer at high monomer conversion, polycondensation becomes the dominant mode by which further increases in molecular weight occur. Temperature had no apparent effect on product molecular weights compared at similar values of monomer conversion. These results led us to the general conclusion that, even though HIC has a structure that largely differs from CALB, the kinetic and mechanistic characteristics of these two enzymes for -caprolactone polymerizations have much in common. The potential use of enzymes for ring-opening of β-butyrolactone and lactide stereoisomers is intriguing because this could offer simple catalysts that provide high levels of stereochemical control during propagation. Unfortunately, the methyl substituents of these monomers rendered them as poor substrates for HIC catalyzed ring-opening polymerizations. Acknowledgment. We are grateful to the members of the NSF I/UCRC for Biocatalysis and Bioprocessing of Macromolecules at the Polytechnic University for their financial support, helpful discussions, and encouragement to pursue this research. We also thank Novozymes for providing us with enzymes. Special thanks to Kenyon College for supporting M.H.’s stay at Polytechnic University during his junior sabbatical leave.
References and Notes (1) Hench, L. L.; Polak, J. M. Science 2002, 295, 1014–1017.
522 Biomacromolecules, Vol. 9, No. 2, 2008 (2) Asuri, P; Karajanagi, S. S.; Kane, R. S.; Dordick, J. S. Small 2007, 3, 50–53. (3) Gross, R. A.; Kumar, A.; Kalra, B Chem. ReV. 2001, 101, 2097–2124. (4) Kobayashi, S.; Uyama, H.; Kimura, S. Chem. ReV. 2001, 101, 3793– 3818. (5) Varma, I. K; Albertsson, A. C.; Rajkhowa, R.; Srivastava, K Prog. Polym. Sci. 2005, 30, 949–981. (6) Henderson, L. A.; Svirkin, Y. Y.; Gross, R. A.; Kaplan, D. L.; Swift, G. Macromolecules 1996, 29, 7759–7766. (7) Mei, Y.; Kumar, A.; Gross, R. A. Macromolecules 2002, 35, 5444– 5448. (8) Mei, Y.; Kumar, A.; Gross, R. Macromolecules 2003, 36, 5530– 5536. (9) Kumar, A.; Gross, R. A. J. Am. Chem. Soc. 2000, 122, 11767–11770. (10) Cordova, A.; Iversen, T.; Hult, K.; Martinelle, M. Polymer 1998, 39, 6519–6524. (11) Thurecht, K. J.; Heise, A.; deGeus, M.; Villarroya, S.; Zhou, J.; Wyatt, M. F.; Howdle, S. M. Macromolecules 2006, 39, 7967–7972. (12) Mee, L. V. D.; Helmich, F.; De Bruijn, R.; Vekemans, J. A. J. M.; Palmans, A. R. A.; Meijer, E. W. Macromolecules 2006, 39, 5021– 5027. (13) Duda, A.; Kowalski, A.; Penczek, S.; Uyama, H.; Kobayashi, S. Macromolecules 2002, 35, 4266–4270. (14) Namekawa, S.; Suda, S.; Uyama, H.; Kobayashi, S Int. J. Biol. Macromol 1999, 25, 145–151. (15) Kumar, A.; Kalra, B.; Dekhterman, A.; Gross, R. A. Macromolecules 2000, 33, 6303–6309. (16) Kolattukudy, P. E. Science 1980, 208, 990–1000.
Hunsen et al. (17) Kolattukudy, P. E. In Lipases; Borgstroem, B., Brockman, H. L., Eds.; Elsevier: Amsterdam, 1984; pp 471–504. (18) Ettinger, W. F.; Thukral, S. K.; Kolattukudy, P. E. Biochemistry 1987, 26, 7883–7892. (19) Longhi, S.; Cambillau, C. Biochim. Biophys. Acta 1999, 1441, 185– 196. (20) Carvalho, C. M. L.; Aires-Barros, M. R.; Cabral, J. M. S. Electron. J. Biotechnol. 1998, 1, 160–173. (21) Carvalho, C. M. L.; Aires-Barros, M. R.; Cabral, J. M. S. Biotechnol. Bioeng. 1999, 66, 17–34. (22) Silva, C.; Carneiro, F.; O’Neill, A.; Fonseca, L. P.; Cabral, J. M. S.; Guebitz, G.; Cavaco-Paulo, A. J. Polym. Sci., Part A: Polym. Chem. 2005, 43, 2448–2450. (23) Yoon, M. Y.; Kellis, J.; Poulose, A. J. AATCC ReV. 2002, 2, 33–36. (24) Hunsen, M.; Azim, A.; Mang, H.; Wallner, S. R.; Ronkvist, A.; Xie, W.; Gross, R. A. Macromolecules 2007, 40, 148–150. (25) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem. 1995, 150, 76. (26) Matsumura, S; Mabuchi, K.; Toshima, K. Macromol. Rapid Commun. 1997, 18, 477–482. (27) Huijser, S.; Staal, B. B. P.; Huang, J.; Duchateau, R.; Koning, C. E. Biomacromolecules 2006, 7, 2465–2469. (28) Nobes, G.; Kazlauskas, R. J.; Marchessault, R. H. Macromolecules 1996, 29, 4829–4833. (29) Gorke, J. T.; Okrasa, K.; Louwagie, A.; Kazlauskas, R. J.; Srienc, F., J. Biotechnol. 2007, 132, 306–313.
BM701269P
