Performance of Selected Microbial Pectinases on Synthetic Monomethyl-esterified Di- and Trigalacturonates*

Two monomethyl esters of α-(1–4)-linked d-galacturonic dimers and three monomethyl esters of α-(1–4)-linked d-galacturonic acid trimers were synthesized chemically and further used as substrates in order to establish the substrate specificity of six different endopolygalacturonases from Aspergillus niger, one exopolygalacturonase from Aspergillus tubingensis, and four selected Erwinia chrysanthemi pectinases; exopolygalacturonan hydrolase X (PehX), exopolygalacturonate lyase X (PelX), exopectate lyase W (PelW), and oligogalacturonan lyase (Ogl). All A. niger endopolygalacturonases (PGs) were unable to hydrolyze the two monomethyldigalacturonates and 2-methyltrigalacturonate, whereas 1-methyltrigalacturonate was only cleaved by PGI, PGII, and PGB albeit at an extremely low rate. The hydrolysis of 3-methyltrigalacturonate into 2-methyldigalacturonate and galacturonate by all endopolygalacturonases demonstrates that these enzymes can accommodate a methylgalacturonate at subsite −2. TheA. tubingensis exopolygalacturonase hydrolyzed the monomethyl-esterified digalacturonates and trigalacturonates although at lower rates than for the corresponding oligogalacturonates. 1-Methyltrigalacturonate was hydrolyzed at the same rate as trigalacturonate which demonstrates that the presence of a methyl ester at the third galacturonic acid from the nonreducing end does not have any effect on the performance of exopolygalacturonase. Of the fourE. chrysanthemi pectinases, Ogl was the only enzyme able to cleave digalacturonate, whereas all four enzymes cleaved trigalacturonate. Ogl does not cleave monomethyl-esterified digalacturonate and trigalacturonate in case the second galacturonic acid residue from the reducing end is methyl-esterified. PehX did not hydrolyze any of the monomethyl-esterified trigalacturonates. The two lyases, PelX and PelW, were both only able to cleave 1-methyltrigalacturonate into Δ4,5-unsaturated 1-methyldigalacturonate and galacturonate.

Fungi from the genus Aspergillus and bacteria from the genus Erwinia produce a wide spectrum of enzymes involved in pectin degradation. Pectin is a heteropolysaccharide with smooth regions of ␣-1-4-linked D-galacturonic acid residues of which approximately 70% of the carboxylic groups are methylesterified. Besides these homogalacturonan parts, the polysaccharide also contains highly branched regions with L-rhamnose and D-galacturonate as the alternating constituents of the backbone structure in which rhamnose can be found to be highly substituted with galactose or long branched arabinose side chains and in which the D-galacturonate moieties can be methyl-esterified, O-acetylated, or substituted with xylose (1). Aspergillus niger was shown to harbor an endopolygalacturonase (PG) 1 gene family consisting of 7 members (2). The PG encoding genes have been cloned and sequenced, and six of them have been overexpressed individually (3)(4)(5)(6). 2 The enzymes have been characterized biochemically using polygalacturonate, pectins with different degree of methyl esterification, and oligogalacturonates with different degree of polymerization. The PGI, -II, -A, and -B appeared to be highly active on the homogalacturonan part of the pectin molecule, whereas the activity of PGC and PGE on this substrate was low (6 -8). Differences were also observed with respect to the influence of the degree of methyl esterification. Whereas PGA and PGB were most active on pectins with a degree of esterification of 22%, the remaining four enzymes preferred non-methyl-esterified polymer substrate. Trigalacturonate was the smallest substrate hydrolyzed by all six PGs. The exopolygalacturonase from Aspergillus tubingensis was shown to hydrolyze D-galacturonate from the nonreducing end of the substrate (9). The enzyme was able to hydrolyze digalacturonate. In a recent study it was demonstrated that exopolygalacturonase is only fully active after de-esterification of the pectin and that the enzyme has an additional xylogalacturonase activity (10).
From the strain Erwinia chrysanthemi 3937, a large number of genes encoding pectinolytic enzymes have been cloned and overexpressed in Escherichia coli (11). The selected enzymes that are included in this study are the exopolygalacturonan hydrolase (PehX), which hydrolyzes digalacturonate from the nonreducing end of polygalacturonate (12), the periplasmic pectate lyase X (PelX) which liberates ⌬4,5-unsaturated digalacturonate from the reducing end of polygalacturonate (12), and the two cytoplasmic enzymes exopectate lyase (PelW) and oligogalacturonan lyase (Ogl) (13,14). For PelW it was demonstrated that the enzyme is active on polygalacturonate and pectin with degree of esterification up to 45% producing ⌬4,5unsaturated digalacturonate (13). Ogl is most active on oligogalacturonates with degree of polymerization less than four, with optimal activity on both saturated and ⌬4,5-unsaturated digalacturonate. Those substrates are degraded into galacturonate and ⌬4,5-unsaturated galacturonate, and 2 molecules of ⌬4,5unsaturated galacturonate, respectively. For ⌬4,5-unsaturated galacturonate, it is known that this compound is not stable and spontaneously converts into 5-keto-4-deoxyuronate (DKI).
Chemical stereoselective synthesis of small oligogalacturonates methyl-esterified in defined positions (15) has opened now the possibility to study the effect of methyl esterification of the substrate on the mode of action of pectinases.

EXPERIMENTAL PROCEDURES
Enzymes-The individual Aspergillus enzymes were purified from the culture fluids of A. niger transformants as described earlier (6 -8).
The protein concentration of the purified enzymes was determined by measuring the absorbance at 280 nm and was calculated according to the method described by Edelhoch (16). The Erwinia enzymes used in this study were expressed in E. coli and isolated as described before (12,13,17).
Removal of Benzyl Groups-To a solution of O-benzylated compound (0.2 mmol) in ethyl acetate (7.5 ml) a mixture of 2-propanol/water/acetic acid: 3:1:1, v/v/v, 46 ml), and Pd(OH) 2 /C (20%, 0.30 g) was added. The reaction mixture was stirred under hydrogen (1 bar) until TLC analysis showed only one spot (3d-f, 7 h, solvent F 1 ; 1h-i, 5 h, solvent F 2 ). The catalyst was removed by filtration and washed with water. The combined filtrates were concentrated under reduced pressure to give, quantitatively, the debenzylated compound, as a colorless solid. In order to get correctly resolved NMR spectra, the crude reaction product was dissolved in water and treated with Dowex 50-H ϩ (0.2 g) for 10 min. Then the resin was filtered off and washed with water. After concentration the residue was dissolved three times in D 2 O (2 ml), and the resulting solution was concentrated under reduced pressure before recording the NMR spectra.
Analytical Methods-Identification of partially methyl-esterified re- action products was performed by high performance anion-exchange chromatography (HPAEC) on a Carbopac PA-100 column (250 ϫ 4 mm, Dionex, Sunnyvale, CA), in 0.02 M sodium acetate buffer, pH 5, using a Dionex BioLC/high performance chromatography system with pulsed amperometric detection (PAD). Samples (50 l, 0.5 mM) were eluted with a 30-min linear gradient of 0.02-0.7 M sodium acetate, pH 5, at a flow of 0.65 ml⅐min Ϫ1 . PAD detection was facilitated by post-column addition of 0.5 M NaOH at a flow of 0.65 ml⅐min Ϫ1 . The purity of the monomethyl-esterified substrates was tested using the same chromatographic conditions.
In order to determine the hydrolysis rate, the reaction mixture was analyzed by high performance anion-exchange chromatography on a Carbopac PA-100 (250 ϫ 4 mm) column equilibrated in 0.2 M NaOH, 0.1 M sodium acetate. Samples were eluted with a 10-min linear gradient of 0.1-0.7 M sodium acetate in 0.2 M NaOH at a flow of 1 ml⅐min Ϫ1 . The eluate was analyzed using PAD detection and, in case of ⌬4,5-unsaturated products, UV-detection at 235 nm. By using this alkaline system the methyl-esterified substrates and reaction products were analyzed as their non-methyl-esterified counterparts. Substrates and saturated products were identified and quantitated using a standard mixture of mono, di-, and trigalacturonate. Unsaturated reaction products were identified by UV detection at 235 nm and a standard mixture of ⌬4,5unsaturated digalacturonate and trigalacturonate. The concentration of the two heterodimers was estimated using digalacturonate as the standard.
Enzyme Incubations-The substrates (50 M) were incubated at 30°C with the Aspergillus enzymes in 50 mM sodium acetate buffer, pH 4.2, or pH 5.0 in case of endopolygalacturonase B. The reaction volume was 0.5 ml. At timed intervals aliquots (70 l) were taken from the incubation mixture, and the enzyme reaction was stopped by raising the pH to 8.5-9.0 by adding 45 l of 2 mM Tris, 50 mM NaOH. Reaction products were analyzed by HPAEC-PAD at pH 12 as described under "Experimental Procedures." The injection volume was 80 l. Initial reaction rates were calculated from the linear increase in time of products and/or decrease of substrate. For A. tubingensis exopolygalacturonase nonlinear progression curves were observed which was due to the strong inhibition of the enzyme by the reaction product, D-galacturonate, as demonstrated earlier (9). In order to minimize the product inhibition, the amount of enzyme used in the incubation mixtures was such that total degradation was less than 3%. For the Erwinia pectinases the reactions were performed at 37°C using 50 M substrate in a total reaction volume of 0.3 ml. Substrates were diluted in 20 mM Tris/HCl, 0.1 mM CaCl 2 , pH 8.0, for Ogl, PelW, and PelX or 50 mM sodium acetate buffer, pH 5.6, for PehX. At timed intervals aliquots (50 l) were taken from the reaction mixture, and the enzyme reaction was stopped by the addition of 10 l of 1.2% acetic acid, for Ogl, PelW, and PelX or 15 l of 2 mM Tris, 50 mM NaOH for PehX. Reaction products were analyzed by HPAEC-PAD at pH 12 as described under "Experimental Procedures." The injection volume was 40 l. Reaction rates were expressed in volume activities as these recombinant enzymes still contained some contaminating host proteins.

RESULTS
Chemically Synthesized Substrates-Before removal of the benzyl groups the synthesized compounds were purified as follows. For 1e, the solvent was evaporated under reduced pressure, and the crude product was dissolved in 30 ml of methanol, and the resulting solution was stirred with 0.5 g of Amberlite 15 resin (H ϩ ). After 15 min the removal of the C-4Ј trimethylsilyl group was complete (TLC, solvent D). The resin was removed by filtration and washed with methanol (3 ϫ 30 ml). The filtrate was concentrated under reduced pressure, and the residue was chromatographed (solvent E) to give compound 1e (0.35 g, 69%) as a syrup. For 1f, the crude reaction mixture was purified by silica gel chromatography (solvent C) to give compound 1f (0.47 g, 88%) as a syrup. For 1g, the crude reaction mixture was purified by silica gel chromatography (solvent C then A 1 ). First 0.208 g of pure compound 1g was eluted as a syrup and then 0.222 g of a mixture (1:4) of 1g and of the ␤-linked anomer (combined yield ϭ 80%; ␣/␤ ϭ 61:39). For 3a, the crude reaction mixture was purified by two silica gel chromatographic steps (1st column, solvent B 2 then B 3 then B 4 ; 2nd column, A 1 then A 2 ). During the second chromatographic step 0.24 g of pure compound 3a, as a syrup, and 0.056 g of a mixture (1:1) of compound 3a and of the ␤-linked isomer (combined yield ϭ 40%; ␣/␤ ϭ 89:11) were eluted. We also recovered 0.270 g (54%) of unreacted acceptor 1c. For 3b, the crude reaction mixture was purified by silica gel chromatography (solvent B 1 then B 2 then B 3 ). First 0.33 g of pure compound 3b was eluted, as a syrup, and finally 0.066 g of a mixture consisting mainly of the ␤-linked anomer (combined yield ϭ 54%; ␣/␤ ϭ 83:17). Unreacted acceptor 1d, 0.160 g (35%), was also recovered. For 3c, the crude reaction mixture was purified by silica gel chromatography (solvent C, then A 2 ). Pure compound 3b (0.27 g, as a syrup) and 0.059 g of a mixture (1:1) of compound 3c and the ␤-linked isomer (combined yield ϭ 45%; ␣/␤ ϭ 95:5). The unreacted acceptor 1e (0.23 g, 50%) was also recovered. After removal of the benzyl groups (see "Experimental Procedures"), the compounds were dissolved in 50 mM sodium acetate buffer, pH 5, and the purity was checked by HPAEC-PAD at pH 5. As for all synthesized compounds the purity was Ͼ95%; they were directly used for the enzymatic degradation studies. The analytical and NMR data of those compounds (1e-i and 3a-f) which have not been published before are shown under "Appendix." A. niger Endopolygalacturonases-The PGs did not hydrolyze 1-and 2-methyldigalacturonate, galactosylgalacturonate, and galacturonylgalactose (results not shown). All six A. niger PGs hydrolyzed 3-methyltrigalacturonate into 2-methyldigalacturonate and galacturonate, although with different efficiencies (Table I). For all enzymes the rate of hydrolysis of 3-methyltrigalacturonate is much lower (6 -18%) than for trigalacturonate, although PGA still retains 38% activity on 3-methyltrigalacturonate. The presence of a methyl group at position 1 or 2 of trigalacturonate influenced the enzyme activity to a much larger extent (Table I). Neither of the PGs was able to hydrolyze 2-methyltrigalacturonate, whereas hydrolysis of 1-methyltrigalacturonate was only observed at a very low rate for PGI, PGII, and PGB.
A. tubingensis Exopolygalacturonase-Exopolygalacturonase is known to hydrolyze di-and trigalacturonate (9), but the two galactose-galacturonate heterodimers were not hydrolyzed by this enzyme (results not shown). All monomethyl-esterified di-and trigalacturonates were hydrolyzed by exopolygalacturonase, although the reaction rates were lower than those for the unmethyl-esterified counterparts, except for 1-methyltrigalacturonate which was hydrolyzed at the same rate as trigalacturonate (Table II). For the degradation of all monomethylesterified oligogalacturonates, nonlinear progression curves were obtained, and the substrates were not fully degraded even after prolonged incubation or in the presence of a large amount of enzyme. This is most likely due to the strong inhibition of the enzyme by the product, galacturonate, as was shown before (9).
E. chrysanthemi Pectinases-None of the four E. chrysan- themi pectinases was able to degrade the two heterodimers of galactose and galacturonate (results not shown). The relative reaction rates for the cleavage of the saturated, unsaturated, and monomethyl-esterified di-and trigalacturonates by Ogl are summarized in Table III and by PelX, PelW, and PehX in Table  IV. Of the four enzymes, Ogl was the only enzyme able to convert the different dimeric substrates. The reaction rate on digalacturonate was almost 5-fold higher than on trigalacturonate. Ogl degraded digalacturonate into galacturonate and DKI. Trigalacturonate was degraded into digalacturonate and DKI. Care was taken to take only first generation events into account. ⌬4,5-Unsaturated digalacturonate and ⌬4,5-unsaturated trigalacturonate were degraded at similar rates as the corresponding saturated compounds. The studies on the degradation of the monomethyl-esterified di-and trigalacturonates demonstrate that the presence of a methyl group at the second galacturonate from the reducing end severely affects the performance of Ogl, whereas a methyl group at the reducing end resulted in a 25-50% reduction of the rate (Table III). In contrast, 3-methyltrigalacturonate was degraded at a 3-fold higher rate than trigalacturonate. The periplasmic PelX released ⌬4,5-unsaturated digalacturonate and galacturonate from trigalacturonate and ⌬4,5-unsaturated digalacturonate and DKI from ⌬4,5-unsaturated trigalacturonate. The presence of the double bond in ⌬4,5unsaturated trigalacturonate severely affected the enzymatic activity (Table IV). The presence of a methyl group at positions 2 or 3 on trigalacturonate also had a severe effect on the activity of PelX. In contrast, only a 3-fold lower reaction rate, compared with trigalacturonate, was found on 1-methyltrigalacturonate.
PelW showed an identical cleavage pattern as PelX. Comparison of the reaction rates of PelW and PelX on trigalacturonate and ⌬4,5-unsaturated trigalacturonate demonstrates that PelW can better accommodate the double bond than PelX (Table IV). Similar to PelX, PelW showed approximately a 3-fold reduction in cleavage rate on 1-methyltrigalacturonate and a low activity on 2-and 3-methyltrigalacturonate as well.
PehX hydrolyzed trigalacturonate into digalacturonate and galacturonate and ⌬4,5-unsaturated trigalacturonate into ⌬4,5-unsaturated digalacturonate and galacturonate at almost equal rates (Table IV). The results in Table IV further demonstrate that PehX is not active on any of the monomethylesterified trigalacturonates. DISCUSSION In this paper we have described the effect of methyl esterification on the activity of 11 pectinases by using monomethylesterified di-and trigalacturonates as the substrates. Until now, studies on the mode of action of pectinases, using well defined substrates, were limited to the use of oligogalacturonates (6 -9, 12, 13). The performance on pectins with various degrees of esterification has been reported (6 -8, 12, 13, 23). The recent study (23) already provided some insight into the tolerance of individual subsites of PGII and exo-PG for methyl esterification of the substrate; however, this approach is not quantitative. The use of well defined monomethyl-esterified diand trigalacturonates allows for an accurate assessment of the tolerance for a methylgalacturonate at the individual subsites around the catalytic site.
A. niger Endopolygalacturonases-The studies using monomethyl-esterified oligogalacturonates clearly demonstrate that the tolerance for binding of a methylgalacturonate at either of the subsites (-1 and ϩ1) adjacent to the catalytic site is absent or low for all PGs. Although the rates for hydrolysis of 3-methyltrigalacturonate are lower than for trigalacturonate (Table I), it is also clear that all PGs are able to bind a methylgalacturonate at subsite Ϫ2. This tolerance for a methyl-esterified residue at subsite Ϫ2 was observed earlier for PGII by analysis of the partially methyl-esterified reaction products after digestion of low methyl-esterified pectin (degree of esterification -31%) (23). The relatively high activity (38%) of PGA on 3-methyltrigalacturonate correlates with the fact that this enzyme shows highest activity on pectin with degree of esterification of 22. 2 In this study only methyl-esterified oligogalacturonates with a degree of polymerization Ͻ4 are used which means that the tolerance or preference for binding of a methylgalacturonate in the case of PGB, which also prefers low esterified pectin, 2 might be determined by a subsite at a further distance from the catalytic site. A complete investigation of the methyl tolerance of all individual subsites has to await the synthesis or purification of monomethyl-esterified oligogalacturonates with a degree of polymerization of 4 -7, thus covering all the subsites.
A. tubingensis Exopolygalacturonase-In contrast to the endopolygalacturonases, for A. tubingensis exopolygalacturonase, the results clearly demonstrate that this enzyme can tolerate a methyl-esterified galacturonate at either subsite Ϫ1 or ϩ1 (Table II). The similar activity of exo-PG on trigalacturonate and 1-methyltrigalacturonate demonstrates further a full tolerance for binding of a methyl-esterified galacturonate at subsite ϩ2. Incubation of a mixture of PGII-derived partially methyl-esterified oligomers with exo-PG followed by analysis of the reaction products by tandem mass spectrometry revealed that this enzyme does not cleave non-esterified galacturonic acid residues from the nonreducing end, if the following residue is methyl-esterified (23). This discrepancy is accounted for by  2 20 36 2-Methyl (GalpA) 2 19 35 (GalpA) 3 136 100 1-Methyl (GalpA) 3 138 100 2-Methyl (GalpA) 3 26 19 3-Methyl (GalpA) 3 25 18  the following. Previously we have shown that exo-PG is strongly inhibited by its product, galacturonate (9). This inhibition also affected the xylogalacturonase activity of the enzyme (10). Furthermore, in the present study hydrolysis of the monomethyl-esterified substrates by exo-PG stopped after 6 -8% conversion, again a demonstration of the potent inhibition of galacturonate. Since the PGII generated oligomer mixture (23) also contained high amounts of galacturonate, it is clear that exo-PG did not act on the oligomers with a methyl ester on the second galacturonate from the nonreducing end. Thus, in pectin degradation, exo-PG will degrade the homogalacturonan parts of the pectin molecule leaving a non-esterified galacturonic acid residue at the nonreducing end.

E. chrysanthemi Pectinases-For
Ogl it is clear that digalacturonate and ⌬4,5-unsaturated digalacturonate are the preferred substrates, which is compatible with an earlier study (13). This study on the monomethyl-esterified oligogalacturonates revealed that the presence of a methyl ester on position 2, occupying subsite Ϫ1, was not tolerated by Ogl (Table III). From the 2-fold reduction in reaction rate for both 1-methyldigalacturonate and 1-methyltrigalacturonate, it can be concluded that Ogl has a moderate tolerance for binding of a methylgalacturonate at subsite ϩ1. The high reaction rate on 3-methyltrigalacturonate cannot be explained directly. The absence of the negative charge may well lead to an increased affinity for subsite position Ϫ2.
PelX attacks polymer substrate from the reducing end. The significant decrease in PelX activity on ⌬4,5-unsaturated trigalacturonate and the absence of activity on 3-methyltrigalacturonate (Table IV) demonstrate that a galacturonate at subsite Ϫ1 is absolutely required for binding of the substrate. Also for subsite ϩ1 an absolute requirement for a carboxylate was observed (see 2-methyl (GalpA) 3 , Table IV), whereas at subsite ϩ2, the presence of a methyl-esterified galacturonate was moderately well tolerated. These data agree well with the previous observations that PelX is composed of 4 subsites (Ϫ2 to ϩ2) and that the activity decreases when the degree of esterification of pectin exceeds 22% (12). Mode of action analysis for PelW suggested the presence of three subsites, Ϫ1, ϩ1, and ϩ2 (13). The absence of subsite Ϫ2 in PelW compared with PelX may be reflected in our observation that PelW is more effective than PelX in cleaving ⌬4,5-unsaturated trigalacturonate (Table IV), i.e. the conformation of the galacturonate at this extremity of the subsite array is less important. However, for effective catalysis, the presence of a free caboxylic group at subsites Ϫ1 and ϩ1 is for PelW as important as for PelX.
Recently, the three-dimensional structure of Erwinia carotovora pectate lyase C (PelC) complexed with pentagalacturonate has been solved. This revealed that for the four galacturonate moieties visible, covering the subsites Ϫ1 to ϩ3, the contact to the enzyme was made via a Ca 2ϩ ion bound to the carboxylate of galacturonate (7). All three lyases used in our study have an absolute requirement for calcium or divalent cations for activ-ity (12)(13)(14). The absolute requirement for a free carboxylic group at subsites Ϫ1 and ϩ1 for PelX and PelW and at subsite Ϫ1 for Ogl suggests that, in analogy with E. carotovora PelC, for these subsites the galacturonate is essential for calcium-or divalent cation-mediated substrate binding.        . We were not able to obtain satisfactory elemental analysis for this very hygroscopic compound.