Biomacromolecules 2001, 2, 958-964
958
Enzymatic Epimerization of Bacterial Mannuronan and of C-6 Oxidized, Galactose-Depleted Guar: A Circular Dichroism and 1H NMR Study Vittorio Crescenzi,* Mariella Dentini, Maria Scala Bernalda, Giancarlo Masci, and Vania Rori Department of Chemistry, University “La Sapienza”, P.le A. Moro 5, 00185 Rome, Italy
Gudmund Skja˚k-Bræk Department of Biotechnology, Norwegian University of Science and Technology, Sem Sælands vei 6-8, 7491 Trondheim, Norway Received March 26, 2001; Revised Manuscript Received June 1, 2001
Attention has been focused on two uronans, namely, mannuronan and galactose-depleted C-6 oxidized guar, the former of microbial origin and the latter of artificial nature, to provide original data on the extent of epimerization they can undergo in dilute aqueous solution using two C-5 mannuronic acid epimerizing enzymes, that is, AlgE-4 and AlgE-6, alone or in admixture. Original circular dichroism data coupled with 1H NMR spectra clearly point out that both uronans can be epimerized, depending on the enzyme or enzyme mixture used, to high levels yielding guluronic-rich alginate samples and guluronic-rich heteropolysaccharides, respectively. Mannuronan and its epimerization products can easily form clear, firm aqueous gels when an excess of HCl is added or when mixed with aqueous CaCl2, respectively. Depleted-guarox does not gel upon addition of excess HCl, while the heterouronan derived from it having a percent of epimerization nearly identical to that of epimerized mannuronan, that is, ca. 70%, can form gel in the presence of Ca(II) only at higher polymer and Ca(II) concentrations. With the latter, heterouronan R-D-galacturonic side groups exert hindrance to “junction zone” formation. Introduction Pursuant to our interest in exploiting natural and artificial uronans having backbones containing D-mannuronic acid residues and, in particular, their enzymatic C-5 epimerization products,1 we have recently reconsidered in detail the case of bacterial mannuronan and of C-6 oxidized, galactosedepleted guar (dep-guarox) using the two enzymes AlgE-4 and AlgE-6 from engineered Escherichia coli. Selected enzymes with mannuronan as substrate introduce alternating mannuronic-guluronic sequences (AlgE-4) or guluronic blocks (AlgE-6).2 It is therefore of interest to ascertain whether both enzymes, alone and in combination, can work equally well on a polymer built up by a mannuronic acid backbone bearing 1,6 linked R-D-galacturonic units (about 1 out of 10 main chain residues, dep-guarox); if so, what are the properties of the ensuing pseudo-alginate? To this end, the circular dichroism (CD) spectra of the two uronans in dilute aqueous solution have been studied as a function of pH, during the course of their partial epimerization using the two epimerases mentioned above, and finally as a function of the amount of added Ca(II) ions. The partially epimerized products have been characterized also in terms of 1H NMR spectra. We wish to report here a summary of the results of such investigations. Experimental Section Guar gum was a Sigma product. High molecular weight mannuronan was a sample produced (in the laboratory of
Prof. G. Skja˚k-Bræk, The University of Trondheim, Norway) using an epimerase-negative mutant (AlgG-) of Pseudomonas fluorescens as described earlier2a for P. aeruginosa. The polymer was deacetylated by treatment with 0.1 M NaOH for 1 h at room temperature, dialyzed, precipitated with 50% ethanol, redissolved in water, submitted to prolonged dialysis, and finally recovered by freeze-drying. Source, purification, and activity of the two epimerases employed, namely, AlgE-4 and AlgE-6, are reported elsewhere.2b-d Regioselective C-6 total oxidation of guar, after partial enzymatic debranching (R-galactosidase), has been performed as previously described in detail;1,3a,b the ensuing sample, dep-guarox, contains about 10% D-galacturonic acid side groups according to 1H NMR analysis. Both mannuronan and dep-guarox were purified by prolonged dialysis against distilled water and recovered by freeze-drying. The degree of C-6 oxidation of the dep-guarox sample, ca. 90%, has been determined by means of potentiometric titration with standard NaOH of a polyelectrolyte solution quantitatively converted to the H+ form by passage through Dowex 50 × 8H+ columns. Circular dichroism spectra were recorded at 25 °C with a Jasco J715-A dichrograph in the wavelength range of 190300 nm with the following setup: bandwidth, 1 nm; time constant, 2 s; scan rate, 50 nm/min; sensitivity, 5 mdeg. Four
10.1021/bm010065+ CCC: $20.00 © 2001 American Chemical Society Published on Web 07/19/2001
Biomacromolecules, Vol. 2, No. 3, 2001 959
Enzymatic Epimerization of Uronans
Figure 1. pH dependence of the circular dichroism spectra of depguarox (A) and of mannuronan (B) in dilute aqueous solution (25 °C). (A) The pH values from (a) to (e) are 2.07, 2.95, 4.65, 5.36, and 6.89. Polymer concentration: 0.2% w/v. (B) The pH values from (a) to (i) are ca. 2 (addition of a slight excess of HCl), 3.02, 3.30, 4.07, 4.38, 4.55, 5.03, 6.51, and 11.18. Polymer concentration: 0.2% w/v. Spectrum a is of a clear, firm gel formed inside the dichrograph cell.
spectra corrected for background were averaged for each sample. Other relevant experimental conditions are indicated in the legends to Figures 1-7. 1H
NMR spectra were recorded, at the temperature indicated in each case in the legends to the figures, using a Bruker AMX 600 spectrometer of the NMR laboratory of the Research Area of Rome, CNR, Montelibretti, Rome, for each NMR spectrum of mannuronan had to be run at a higher temperature than for all other samples because of the much higher molecular weight of the biopolymer and the very viscous nature of its aqueous solutions. In the spectra of Figures 3 and 5, M and G denote internal GulA and ManA residues, the numbers denote which proton is causing the signal, and the nonunderlined letters refer to neighboring residues. 1H
To estimate the degree of epimerization from NMR spectra of mannuronan treated with AlgE-4, the ratio between the area of the guluronic acid peak (at around 5.1 ppm) and the area of all peaks in the range of 4.7-4.9 ppm (anomeric region) has been considered. For epimerizations carried out with AlgE-6 and with a mixture of AlgE-4 and AlgE-6, the ratio of the areas of peaks at about 5.1 ppm and the area of all peaks in the range of 4.44-4.84 ppm has been calculated. Similarly, in the case of dep-guarox (Figure 5), the degree of epimerization has been evaluated as the ratio of the areas
Figure 2. Time course of the partial epimerization of mannuronan in dilute aqueous solution as monitored by circular dichroism at 40 °C (50 mM Tris/HCl buffer, pH ) 6.9; 50 mM NaClO4; 2 mM CaCl2; polymer concentration, 0.1% w/v). Enzyme employed: (A) AlgE-4, 0.005% w/v; (B) AlgE-6, 0.01% w/v; (C) AlgE-4 (0.005% w/v) and AlgE-6 (0.01% w/v) together.
of peaks at around 5.1 ppm (subtracted of the area of the GalA peak for dep-guarox reported in Figure 5a) and the areas of all peaks in the 4.44-4.48 ppm region.4 In all cases, therefore, a deconvolution of the anomeric peaks was not necessary. Results and Discussion pH Dependence of the CD Spectra of Aqueous Mannuronan and Dep-Guarox. The CD spectra of mannuronan and of dep-guarox recorded in a wide pH range are reported
960
Biomacromolecules, Vol. 2, No. 3, 2001
Crescenzi et al.
Figure 3. 1H NMR spectra (0.01 M NaCl in D2O) of (a) mannuronan (90 °C) and of mannuronan epimerized for 48 h (see Figure 2) using (b) AlgE-4 (70 °C), (c) AlgE-6 (70 °C), and (d) mixed AlgE-4 and AlgE-6 (64 °C). M and G denote internal GulA and ManA residues, the numbers denote which proton is causing the signal, and the nonunderlined letters refer to the neighboring residues. The estimated degrees of epimerization (see Experimental Section) are (b) 40%, (c) ca. 70%, and (d) ca. 70%.
in Figure 1. For pH values in the approximate interval of 3-11, both sets of spectra clearly exhibit an isodichroic point at around 227 nm indicative of a simple pH-dependent equilibrium of the carboxyl chromophores between two forms only, that is, ionized and protonated. However, upon addition of a slight excess of HCl the solution of mannuronan suddenly sets to a clear, rigid gel and the spectrum undergoes a relatively large shift. The phenomenon does not occur with dep-guarox. Thus, bacterial mannuronan, but not dep-guarox, is a good gelling agent in acid media, the clear wall to wall solidlike phase formation being accompanied by a marked spectral perturbation, probably a consequence of the establishment, in a cooperative fashion, of an extensive array of interchain hydrogen bonds in which a number of carboxylic groups would be engaged. Epimerizations Followed via CD and 1H NMR Measurements. The CD spectra of mannuronan recorded at various times after addition of AlgE-4, of AlgE-6, and of a mixture of the two enzymes are reported in parts A, B, and C of Figure 2, respectively. Spectral changes consequent to epimerization are clear in all instances and seem particularly evident in the case of AlgE-6 and of the mixed enzymes.
It must be pointed out, however, that the initial epimerization velocity of AlgE-4 is distinctly higher than that of AlgE-6, and this explains why the “0 h” spectrum of Figure 2A (recorded in practice some 5 min after having added the enzyme solution under stirring) is different from the corresponding “0 h” spectra obtained after addition of AlgE-6 or of the enzyme mixture. In qualitative terms, the trend of the spectra with time reflects, as expected, the transformation of D-mannuronic acid into L-guluronic acid residues along the chains, and it reflects also the interactions of L-guluronic acid residues with calcium ions present in the reaction solution (concentration ratio Ca(II) ions/polymer equal to 0.7).5 However, there are obvious differences among the three sets of CD spectra of Figure 2: in particular, the results obtained with the mixed enzymes do not appear as a simple combination of what is observed with AlgE-4 and AlgE-6 separately. From the 1H NMR spectra of Figure 3, it can be estimated that after 48 h the degree of epimerization of mannuronan is about 40% having used AlgE-4 and between 70 and 80% having used AlgE-6 or the mixture of the two enzymes. The CD spectra of dep-guarox recorded at various times of enzymatic treatment for experimental conditions nearly
Enzymatic Epimerization of Uronans
Biomacromolecules, Vol. 2, No. 3, 2001 961
Figure 4. Time course of the partial epimerization of dep-guarox in dilute aqueous solution as monitored by circular dichroism at 40 °C (50 mM Tris/HCl buffer, pH ) 6.9; 50 mM NaClO4; 2 mM CaCl2; polymer concentration, 0.1% w/v). Enzyme employed: (A) AlgE-4, 0.005% w/v; (B) AlgE-6, 0.01% w/v; (C) AlgE-4 (0.005% w/v) and AlgE-6 (0.01% w/v) together.
962
Biomacromolecules, Vol. 2, No. 3, 2001
Crescenzi et al.
Figure 5. 1H NMR spectra (0.01 M NaCl in D2O) of (a) dep-guarox (70 °C) and of dep-guarox epimerized for 48 h (see Figure 4) using (b) AlgE-4 (70 °C), (c) AlgE-6 (70 °C), and (d) mixed AlgE-4 and AlgE-6 (70 °C). M and G denote internal GulA and ManA residues, the numbers denote which proton is causing the signal, and the nonunderlined letters refer to the neighboring residues. The estimated degrees of epimerization are (b) 29%, (c) 40%, and (d) 70%.
identical to those used for mannuronan (see Figure 2) are collected in Figure 4. Comparison of Figures 4 and 2 shows a different trend of the spectra with increasing epimerization, but again, one observes that spectra recorded using the mixed enzymes do not appear to be a linear combination of those obtained with the single epimerases. From the 1H NMR spectra of Figure 5, it can be estimated (see Experimental Section) that the maximum extent of depguarox epimerization is ca. 30% with AlgE-4, ca. 40% with AlgE-6, and ca. 70% with the combined enzymes. The latter figure is noteworthy inasmuch as it parallels that found with the natural substrate, mannuronan, indicating that D-galacturonic side chains in dep-guarox do not hinder the epimerization process of backbone residues at all. Influence of Added Ca(II) Ions on the CD Spectra of Epimerized Mannuoronan and Dep-Guarox. Data of Figure 6 show that with partially epimerized dep-guarox the change in the CD spectrum brought about by the addition of Ca(II) ions is relatively modest (as in the case of mannuronan, spectra not shown) while to the contrary, with partially epimerized mannuronan there are major variations in circular dichroism (Figure 7). Spectral changes moreover depend on which enzyme (or enzyme mixture) has been used. In fact, the set of spectra in
part A of Figure 7 pertaining to mannuronan partially epimerized using AlgE-4 are different from those in parts B and C relative to mannuronan epimerized with AlgE-6 or with the enzyme mixture, respectively. Evidently, the selective Ca(II) binding properties of block sequences of Lguluronic acid residues introduced along mannuronan chains by AlgE-6 and by AlgE-4 + AlgE-6 find a counterpart in a peculiar, marked change in CD features which appear particularly evident in the case of the epimerization product of AlgE-6. In addition, separate experiments performed using polymer concentrations about 0.3% w/v or higher and R values of 4 (see legend to Figure 7) demonstrate that AlgE-6 epimerized mannuronan can form clear, stable gels typical of L-guluronic acid rich alginates, as expected. With dep-guarox epimerized using the enzyme mixture (epimerization of about 70%), Ca(II)-induced gelation takes place only for higher polymer and Ca(II) concentrations, for example, 2.5% w/v and R ) 4. Concluding Remarks From the results reported above, it can be concluded that the combined use of CD and NMR measurements appears particularly useful in monitoring enzymatic epimerization processes, at least for cases considered in this work.
Enzymatic Epimerization of Uronans
Figure 6. Dependence on Ca(II) ion concentration of the CD spectrum of dep-guarox partially epimerized (48 h, see Figure 4) with (A) AlgE-4, (B) AlgE-6, and (C) mixed enzymes. The polymer concentration is 0.02% w/v. R is the concentration ratio (Ca(II) ions/ polymer) in equivalents.
Our present data confirm the ability of C-5 mannuronic epimerases AlgE-4 and AlgE-6 to deal with non-natural substrates containing β-(1,4) linked D-mannuronic acid residues either flanked by other types of acidic residues (as for instance in the case of regioselectively C-6 oxidized Konjac mannan1) or bearing relatively low amounts of charged side groups in dep-guarox. With the latter, the C-5 epimerization process has reached levels (about 70%) comparable to those of mannuronan for identical experimental conditions. Nevertheless, with 70% epimerized dep-guarox for Ca(II)-induced gelation higher
Biomacromolecules, Vol. 2, No. 3, 2001 963
Figure 7. Dependence on Ca(II) ion concentration of the CD spectrum of mannuronan partially epimerized (48 h, see Figure 4) with (A) AlgE-4, (B) AlgE-6, and (C) mixed enzymes. The polymer concentration is 0.02% w/v. R is the concentration ratio (Ca(II) ions/ polymer) in equivalents.
polymer and Ca(II) concentrations are needed (polymer concentrated about 2.5% w/v and R ) 4). This may be traced to a hindering of the typical “eggbox” structure formation exerted by D-galacturonic acid side groups. The inability of dep-guarox to undergo acid-induced gelation may have the same origin (i.e., hindrance to partial chain pairing). However, there is a large difference in average molecular weight between polycarboxylates considered, that is, ca. 1000 kDa for mannuronan and ca. 100 kDa for depguarox (values estimated through viscosity data, not shown), which may conjure to widen differences in their solution/ gelling properties.
964
Biomacromolecules, Vol. 2, No. 3, 2001
Studies are in progress in our laboratory on the solution behavior of epimerization products, obtainable using also different epimerases, of dep-guarox in which the amount of side groups has been further reduced or totally eliminated. More important, a topic for further studies concerns a better understanding of the mechanism of action of epimerases when in admixture; CD data presented here can only incline to suspect a non-“additive” behavior. Acknowledgment. This work has been carried out with financial support of the European Union (Contract QLK3CT-1999-00034) and of the Italian Ministry for the Universities and Scientific and Technological Research, MURST (Contract “cofinanziato-1999”). The expert assistance of Professor A. L. Segre in running and interpreting the NMR spectra is gratefully acknowledged.
Crescenzi et al.
References and Notes (1) Crescenzi, V.; Hartmann, M.; de Nooy, A. E. J.; Rori, V.; Masci, G.; Skja˚k-Bræk, G. Biomacromolecules 2000, 1 (3), 360. (2) (a) Ertesvag, H.; Skja˚k-Bræk, G. Methods Biotechnol. 1999, 10, 71. (b) Ertesvag, H.; Doseth, B.; Larsen, B.; Skja˚k-Bræk, G.; Valla, S. J. Bacteriol. 1994, 176, 2846. (c) Ertesvag, H.; Hoidal, H. K.; Skja˚kBræk, G.; Valla, S. J. Biol. Chem. 1998, 273, 30927. (d) Svanem, B. J. G.; Skja˚k-Bræk, G.; Ertesvag, H.; Valla, S. J. Bacteriol. 1999, 181, 68. (3) (a) de Nooy, A. E. J.; Besemer, A. C.; van Bekkum, H. Carbohydr. Res. 1995, 269, 89. (b) de Nooy, A. E. J.; Rori, V.; Masci, G.; Dentini, M.; Crescenzi, V. Carbohydr. Res. 2000, 324, 116. (4) Grasdalen, H.; Larsen, B.; Smidsrød, O. Carbohydr. Res. 1979, 68, 23. (5) Morris, E. R.; Rees, D. A.; Thom, D. Carbohydr. Res. 1980, 81, 305.
BM010065+
