Search tips
Search criteria 


Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. 2009 December; 75(23): 7343–7349.
Published online 2009 October 9. doi:  10.1128/AEM.01010-09
PMCID: PMC2786414

Thermal Stabilization of Erwinia chrysanthemi Pectin Methylesterase A for Application in a Sugar Beet Pulp Biorefinery[down-pointing small open triangle]


Directed evolution approaches were used to construct a thermally stabilized variant of Erwinia chrysanthemi pectin methylesterase A. The final evolved enzyme has four amino acid substitutions that together confer a Tm value that is approximately 11°C greater than that of the wild-type enzyme, while maintaining near-wild-type kinetic properties. The specific activity, with saturating substrate, of the thermally stabilized enzyme is greater than that of the wild-type enzyme when both are operating at their respective optimal temperatures, 60°C and 50°C. The engineered enzyme may be useful for saccharification of biomass, such as sugar beet pulp, with relatively high pectin content. In particular, the engineered enzyme is able to function in biomass up to temperatures of 65°C without significant loss of activity. Specifically, the thermally stabilized enzyme facilitates the saccharification of sugar beet pulp by the commercial pectinase preparation Pectinex Ultra SPL. Added pectin methylesterase increases the initial rate of sugar production by approximately 50%.

Pectin is a heterogenous structural polysaccharide found in plant primary cell walls. Pectin helps to connect and cross-link other cell wall polysaccharides, such as cellulose and hemicellulose, to contribute to cell wall rigidity. The pectin backbone is comprised of α-(1,4)-linked galacturonic acid (GalA) subunits. In addition to the “smooth” regions of homogalacturonans, there are variable proportions of “hairy” regions consisting of α-(1,5)-linked arabinans and/or β-(1,4)-linked galactans as well as other neutral sugars. Some of the GalA subunits carry methoxyl and acetyl esters at C-6 and C-2/C-3, respectively. The degree of C-6 methylesterification, in particular, influences the rheological properties of the polymer (18, 19, 24).

Pectin methylesterases (PMEs; EC catalyze the demethylesterification of GalA C-6 producing methanol, protons, and polygalacturonate. This reaction is significant in a number of contexts. In muro, the activity of plant PMEs helps control cell wall rigidity and plays a major role in pectin remodeling related to cell wall growth and processes such as fruit ripening (15). In the case of bacterial and fungal phytopathogens, PMEs are virulence factors that are necessary for pathogen invasion and spread through plant tissues (2, 3, 23). PMEs along with other pectinolytic enzymes are widely used in the food and beverage industries and paper and fiber industries, among others (9).

PMEs work in concert with other pectinolytic enzymes, pectate lyases (EC, and pectate glycohydrolases (EC, among others, to depolymerize pectin. Highly esterified pectin is largely resistant to depolymerization (1). Shevchik and colleagues demonstrated that pretreatment of purified sugar beet pectin with the Erwinia chrysanthemi PMEA resulted in a 10- to 20-fold enhancement in the catalytic rate of E. chrysanthemi pectate lyases PELA, PELB, PELC, PELD, and PELL relative to that of the untreated substrate (21). Similarly, Christgau and colleagues found that depolymerization of purified apple pectin by pectate glycohydrolase from Aspergillus aculeatus was dependent on added PME (5).

E. chrysanthemi produces at least two PMEs. PMEA is a 342-amino-acid secreted protein, and the 433-amino-acid PMEB is bound to the outer membrane (12, 20). PMEA is a novel aspartate-esterase that folds into a right-handed parallel β-helix, similar to other pectinolytic enzymes, such as pectate lyases and polygalacturonases (8, 10). The enzyme is active over a broad pH range (5-9) and has optimal activity around 50°C (11, 16).

Sugar beet pulp is the by-product of sucrose production from the tap root of Beta vulgaris. Sugar beet pulp is rich in pectin, hemicellulose, and cellulose and relatively low in lignin content. It exits the sugar refinery as heated (ca. 60°C) thin slices approximately 75% water by weight. These and other considerations make sugar beet pulp an attractive biomass target for enzymatic saccharification and subsequent conversion of sugars to value-added products.

In order to create a PME suited to the saccharification of sugar beet pulp, we employed directed evolution approaches to engineer a variant of E. chrysanthemi PMEA that would function at 60°C in sugar beet pulp. Here, we report the development of a thermostabilized PMEA variant with four amino acid substitutions that demonstrates efficacy in the saccharification of sugar beet pulp.



Ruthenium red, citrus pectin, apple pectin, polygalacturonic acid, Hansenula alcohol oxidase, sodium ampicillin, l-arabinose, and Pectinex Ultra SPL were obtained from Sigma. Fluoral-P was obtained from Acros Organics. Sugar beet pulp was obtained from Michigan Sugar. All other reagents were from Fisher unless noted otherwise.

Cloning pmeA.

pmeA was amplified from E. chrysanthemi strain 3937 using 5′-GCCATGGATATGTTAAAAACGATCTCTG-3′ forward and 5′-GCAATTCGTCAGGGTAATGTC-3′ reverse primers. The forward primer contains an NcoI site (CCATGG) to coincide with the start site of the open reading frame. PCR amplification conditions were as follows: 10 min at 94°C, then 1 min at 94°C, 1 min at 57°C, and 90 s at 72°C, for 30 cycles, and 10 min at 72°C on a Stratagene Robocyler. The PCR product was purified using the Qiaquick PCR purification kit (Qiagen). The cloning used the pBAD/TOPO expression kit (Invitrogen). The PCR product was ligated into the pBAD/Thio-TOPO vector, and the recombinant vector was transformed into chemically competent Escherichia coli Top10 cells. Subsequently, the 5′ His-patch-thioredoxin fusion sequence between the araBAD promoter and the start of the pmeA sequence was removed by NcoI digestion and religation to create pBAD-pmeA.

Error-prone PCR.

The GeneMorph II EZClone domain mutagenesis kit (Stratagene) was used. Briefly, 50 ng of pBAD-pmeA plasmid was used as a template with 5′-AGGTATACATACCCATGGATATG-3′ forward and 5′-GTCAGTTGCCGGCGGTCT-3′ reverse primers. PCR amplification conditions were as follows: 2 min at 95°C followed by 1 min at 94°C, 1 min at 55°C, and 2 min at 72°C, for 30 cycles, and 10 min at 72°C. This error-prone PCR product created a truncated sequence of 1,034 bp (missing 67 bp at the 3′ end). This PCR product was purified with the Strataprep PCR purification kit (Stratagene) and used as a megaprimer to recreate the entire expression vector. Five hundred nanograms of megaprimer and 50 ng of recombinant pBAD-pmeA vector were cycled under the following conditions: 1 min at 95°C followed by 50 s at 95°C and 50 s at 60°C and 10 min at 68°C, for 25 cycles. The mutagenized pBAD-pmeA vector was then transformed into XL10-Gold ultracompetent E. coli (Stratagene).

Site-saturation mutagenesis.

Oligonucleotides with 64-fold degeneracy at the target codon were used with the QuikChange multisite-directed mutagenesis kit (Stratagene) and the appropriate pBAD-pmeA template to explore all amino acid substitutions at the targeted residue.

Recombinant protein expression.

E. coli cells harboring pBAD-pmeA were transferred by toothpick from colonies growing on Luria broth (LB) agar plates with 100 μg/ml ampicillin (LB-amp) into 200 μl of LB-amp in 96-well plates (Costar). The cultures were incubated at 37°C in a humidified chamber with shaking overnight. Following overnight growth, the cultures were diluted 1:20 into fresh LB-amp and grown as described above until the mean A595 (measured on a Tecan Genios plate reader) of the wells was between 0.5 and 0.8. The cultures were then diluted 1:7 into fresh LB-amp, 0.5% l-arabinose and incubated as described above. When the mean A595 reached between 0.3 and 0.4, the cultures were centrifuged at 4,000 × g for 15 min at 4°C to collect the induced supernatants.

Ruthenium red agar diffusion assay.

The ruthenium red agar diffusion assay was adapted from Downie et al. (7). Assay medium [10.5 g K2HPO4, 4.5 g KH2PO4, 1 g (NH4)2SO4, 15 g Bacto agar, and 2.5 g citrus pectin per liter] was brought to a boil with vigorous stirring until all of the agar and pectin went into solution and was then poured in 150- by 15-mm VWR petri dishes. Small volumes (2 to 5 μl) of induced supernatant from 96-well plates were pipetted onto the surface of the assay medium in the corresponding array pattern. The plates were incubated at 37°C for 15 to 60 min. The plates were then washed with 15 ml of tap water at least two times. Then, 3 ml of 0.5% (wt/vol) ruthenium red was pipetted onto the plates and spread evenly. After 1 min exposure to the dye, the plates were washed with tap water and scored. Ruthenium red binds preferably to the anionic polygalacturonic acid rather than to the methyl-esterified pectin.

Continuous fluorometric PME assay.

The method of Wojciechowski and Fall was adapted with slight modifications (26). Dilutions of induced supernatants or purified enzymes in 15 μl were added to 85 μl of assay reagent (100 mM KHPO4 [pH 7.5], 150 mM NaCl, 0.5% [wt/vol] apple pectin, 10 μg/ml Fluoral-P, 10 U/ml Hansenula alcohol oxidase) in a black-walled 96-well plate (Greiner). The assay mixture was incubated at 30°C, while taking fluorescence readings (405 nm excitation, 535 nm emission) every 2 minutes (Tecan Genios). PME activity (the rate of methanol production) was calculated by subtracting any spontaneous reaction and recording the slope of the linear regression line during the linear portion of the reaction. The slopes associated with known, added amounts of methanol were used to construct a standard curve to relate the slopes of the assay mixtures to the amount of methanol produced.

PME purification.

The method of Laurent and colleagues (11) was used with some modification. Cross-linked polygalacturonate affinity matrix was prepared as described previously (11). Cultures (100 ml LB-amp) of E. coli Top10 harboring the appropriate pBAD-pmeA plasmid were grown at 37°C to early log phase and then induced by addition of l-arabinose to a final concentration of 0.2% (wt/vol). Following arabinose addition, the cultures were incubated at 22 to 24°C for 16 to 18 h. Culture supernatants were clarified by two successive centrifugations at 4,000 × g for 20 min at 4°C. Supernatants were brought to 45% saturation with (NH4)2SO4, stirred for 1 hour at 4°C, and centrifuged at 12,000 × g for 20 min at 4°C. The precipitates were discarded, and the remaining supernatants were brought to 80% saturation with (NH4)2SO4, stirred for 1 hour at 4°C, and centrifuged at 12,000 × g for 20 min at 4°C. The precipitates were resuspended in the minimal volume of 20 mM sodium succinate pH 6.0 (buffer A). To concentrate the samples and remove any residual (NH4)2SO4, the samples were washed three times with buffer A and brought to a final volume of 1.5 ml by ultrafiltration (Pierce). Affinity chromatography columns were prepared by resuspending the cross-linked polygalacturonate in buffer A and filling columns (Pierce) with an ~3-ml bed volume of matrix. The columns were washed with three column volumes of buffer A, and then loaded with the concentrated, desalted PME sample. Unbound material was removed by three washes with 3 ml buffer A. Bound PME was eluted with 4 ml of 0.3 M NaCl-0.2 M KHPO4 (pH 7.5) buffer. The eluted fraction was desalted and concentrated as described above back into buffer A for storage at 4°C. Protein concentration was determined by the method of Bradford (4), and purity was assessed by visualizing proteins after sodium dodecyl sulfate-polyacrylamide gel electrophoresis separation and staining with the GelCode blue safe stain reagent (Pierce).

Determination of Tm.

Induced supernatants were diluted fivefold in LB, and 50 μl was transferred to rows of 96-well Thermowell plates (Costar). The samples were heated on the thermal gradient block of a Robocycler (Stratagene) at the indicated temperature (see Fig. Fig.4)4) for 20 min. Fifteen-microliter aliquots of the heated samples were assayed for residual activity with the saturating substrate by the continuous fluorometric PME assay.

FIG. 4.
Thermostability of PME variants. Induced supernatants were heated for 20 min at the indicated temperature and were then assayed for initial activity in the AO-coupled fluorescent assay. The initial velocities with the saturating substrate were normalized ...

Temperature dependence of activity.

Purified enzymes were added to chilled assay reagent (0.5% apple pectin, 100 mM NaCl, 20 mM succinate; pH 6.0) to a final concentration of 0.1 μg/ml. The enzyme reaction mixture was brought to the indicated temperature (see Fig. Fig.5A)5A) for 15 min followed by 2 min at 75°C to inactivate the enzyme. Aliquots of this reaction mixture were diluted in buffer A and assayed for methanol by the fluorometric alcohol oxidase assay described above but without added pectin substrate.

FIG. 5.
Characterization of the thermal dependence of JL25 PME activity. (A) The relative activity was calculated by subtracting any spontaneous reaction from each enzyme-catalyzed reaction at the corresponding temperature and normalizing each reaction to the ...

Kinetics of thermal inactivation.

Purified enzymes were added to 20 mM succinate (pH 6.0) (buffer B) prewarmed to the indicated temperature (see Fig. Fig.5B)5B) to a final concentration of 2 μg/ml. Aliquots (15 μl) were removed at the indicated times directly into 30 μl of 0.15 M NaCl-0.1 M KHPO4 (pH 7.5) at 4°C and kept on ice until assayed for residual activity at 30°C by the continuous fluorometric PME assay.

Determination of enzyme kinetic constants.

Purified enzymes (15 μl of 3 μg/ml) in buffer B were added to 85 μl of the continuous fluorometric PME assay reagent as described above but with pectin concentrations varying between 0.0022% and 0.5%. The rate of methanol production was determined as described above. Data were fitted to Michaelis-Menten kinetics by the use of SigmaPlot software.

Digestion of sugar beet pulp.

Fifty grams of beet pulp in a total volume of 100 ml of 50 mM MOPS (morpholinepropanesulfonic acid; pH 6.5) was digested at 55°C with constant rotation with 260 U/ml Pectinex Ultra SPL with and without added JL25 PME (5 ng/ml). Samples were removed at the times indicated below, filtered through a 0.45-um Millex HA filter (Millipore), and assayed for total reducing sugar concentration by the dinitrosalicylic acid assay (13).


Directed evolution approaches were employed to construct a thermally stabilized PME. Screening of error-prone, PCR-generated libraries derived from the Erwinia chrysanthemi pmeA sequence identified several amino acid residues where thermostabilizing substitutions were found. Saturation mutagenesis at these sites and subsequent combination of the substitutions into one molecule resulted in a PME with an elevated Tm (62.5°C versus 51°C) relative to that of the parental enzyme. Figure Figure11 summarizes the amino acid substitutions and increases in thermal stability as additional substitutions were combined.

FIG. 1.
Summary of the thermally stabilized PME variants after each generation of mutagenesis. The bars indicate the temperature at which 50% of the initial activity was retained after 20 min of heating. The genotypes of the variants are indicated below ...

In order to identify point mutations that lead to thermal stabilization, we developed a rapid screening assay for residual activity following heating. The assay was adapted from the ruthenium red agar gel diffusion assay of Downie and colleagues (7). Low volumes of heated culture supernatant from library clones were pipetted in an array pattern corresponding to 96-well plates onto agar plates containing citrus pectin. The reaction plates were incubated at 37°C, and PME activity was detected by the formation of pink-stained areas after brief exposure of the plates to a solution of ruthenium red and rinsing with water. Figure Figure22 contains images of typical assay results of a library. Note from Fig. Fig.2A2A that by assaying unheated culture supernatants, one can readily assess the fraction of the library clones with mutations that lead to loss of activity (null mutation). The arrow in Fig. Fig.2B2B points to a clone harboring a candidate thermostabilizing mutation. In addition to monitoring the null mutation rates of the libraries, a quantitative fluorometric assay (26) for PME activity was used to assess library diversity. Figure Figure33 depicts the distributions of enzyme activities from a mutagenized library compared to unmutagenized sibling clones. The normalized distribution of unmutagenized sibling clones had a mean of 1.0 with a standard deviation of 0.05 and a range of relative activity from 0.86 to 1.23. The normalized distribution of mutagenized library clones had a mean of 1.0 with a standard deviation of 0.67 and a range of relative activity from 0 to 1.90. This establishes that the assay itself has relatively low variability and that there is a rich diversity of activities in the mutagenized library. A total of 3,440 library clones were screened for residual activity following heating at 54°C for 20 min. From these, we initially found four different single amino acid substitutions that were confirmed to impart enhanced thermal stability (Fig. (Fig.1,1, generation 1).

FIG. 2.
Results of ruthenium red agar diffusion assay for PME activity. (A) Assay of 3 μl of unheated, induced supernatants from an error-prone PCR library. (B) Assay of 3 μl of induced supernatants heated for 20 min at 53°C. The arrow ...
FIG. 3.
Coupled alcohol oxidase fluorescent assay of induced supernatants from unmutagenized clones and clones from error-prone PCR library. Assay of 15 μl of induced supernatants from unmutagenized, wild-type clones (circles) and mutagenized library ...

Beginning with the CC23 background (R253K) and doing site-saturation mutagenesis at codon 273 and screening for residual activity after heating to 58°C, we found six active clones. Five recapitulated the substitution found in the first-generation mutant CC7 (T273R), and one contained a T273K substitution. These clones are referred to as JL11 and JL10, respectively. Site-saturation mutagenesis of codon 251 from the JL10 and JL11 backgrounds and screening after heating to 59°C yielded three clones with cysteine substitutions (JL18) and one clone with an alanine substitution (JL19), respectively. Next, we screened over 900 clones at 60°C following site-saturation mutagenesis at codon 158 and did not find any clones with further gains in thermostability.

Taking the JL18 background (R253K, T273K, V251C) and preparing an error-prone PCR library (14.7% null rate), we screened over 1,000 clones for residual activity after heating to 62°C. We found one clone with activity, JL24. JL24 has two additional substitutions, A16G and S228F. We constructed mutants with each substitution added singly to the JL19 background. JL19 with the S228F substitution, designated JL25 (R253K, T273K, V251C, S228F), demonstrated residual activity after heating to 62°C, while JL19 with the A16G substitution had a melting temperature indistinguishable from that of JL19. From this, we concluded that the S228F substitution was solely responsible for the increase in thermostability. The combined effect of the four stabilizing substitutions in JL25 resulted in an enzyme with a melting temperature of 62.4 ± 0.1°C compared to the wild-type enzyme's melting temperature of 51.3 ± 0.3°C. The residual activity of the various mutant proteins after heating to the temperature indicated for 20 min is shown in Fig. Fig.44.

Returning to the Y158H substitution found in the first-generation mutagenesis, we did site-directed mutagenesis to specifically introduce the Y158H substitution into the JL25 background. Addition of the Y158H substitution increased thermostability only marginally, by 0.5°C or less (data not shown).

To determine the relative temperature dependence of activities and the rates of thermal inactivation, we purified the parental PME and JL25 mutant enzymes. Figure Figure5A5A shows the relative rates of catalysis by the parental enzyme and the thermostabilized JL25 variant with saturating substrate at various relevant temperatures. The rates of thermal inactivation of the parental enzyme and the JL25 variant (Fig. (Fig.5B)5B) were determined by heating aliquots of the proteins in the absence of substrate at the indicated temperature and time and then assaying the residual activity. Under these conditions at 62°C, JL25 showed no diminution in activity, whereas at 65°C, JL25 had a half-life of 13.9 ± 1.6 min. At 55°C, the parental enzyme had a half-life of 5.7 ± 1.2 min. This is in good agreement with the findings of Laurent et al., who reported a half-life of 5 min at 55°C (12).

The kinetic constants were derived for the wild-type enzyme and the JL25 enzyme. The enzymes obeyed Michaelis-Menten kinetics, and the Km value for the wild-type enzyme, 0.044 ± 0.010% pectin, is in excellent agreement with that reported by Laurent and colleagues and Fries et al., 0.03 to 0.05% (8, 11). However, the catalytic constant we derived, 688 ± 68 s−1, is greater than the value reported by Fries et al., 450 s−1. The JL25 enzyme had Km and kcat values of 0.073 ± 0.023% and 655 ± 104 s−1, respectively.

A goal of this work was to develop enzymes that facilitate the saccharification of sugar beet pulp. As the spent pulp exits the refinery, it is still at a relatively elevated temperature, approximately 60°C. Table Table11 shows the kinetics of production of reducing sugars by a commercial preparation of pectinase (Pectinex Ultra SPL) with and without added JL25 PME. Supplementing Pectinex Ultra SPL with the JL25 PME facilitated the saccharification of the beet pulp polysaccharides. Over the first 12 h of the reaction, the reaction with the PME was approximately 50% faster than the nonsupplemented reaction. Most, but not all, of the additional sugar released in the reaction with the PME was galacturonic acid (data not shown).

Time course of sugars produced from sugar beet pulp


Directed evolution approaches were used to construct a thermally stabilized variant of Erwinia chrysanthemi PMEA. The final evolved enzyme, JL25, has a Tm that is approximately 11°C greater than that of the wild-type enzyme and has a catalytic constant that is not much, if any, diminished relative to that of the wild type. Furthermore, JL25 has a specific activity slightly greater than that of the wild-type enzyme operating with saturating substrate at their optimal temperatures of 60°C and 50°C, respectively. The JL25 variant has optimal catalytic activity between 60° and 65°C.

We employed a convenient expression and screening procedure to find PME mutants with desired phenotypes. Using E. coli as an expression host was advantageous, as it does not express any background PME activity yet recognizes and acts on the native secretion signals of the E. chrysanthemi PME. Working with a secreted protein obviates the need for cell lysis before assaying. This eliminates an extra step that can lead to additional variability in screening assay results. The rapid and inexpensive ruthenium red agar diffusion assay allowed us to screen many candidate clones without expensive equipment. Finally, the coupled PME-alcohol oxidase (AO) fluorescent assay provided a reliable, quantitative assay to confirm the phenotypes of candidate mutants. The tight distribution (Fig. (Fig.3)3) of activities from sibling clones in this assay makes it suitable for other screens, such as those for increases in catalytic activity. Øbro and colleagues recently published a novel method for screening PME variants by the use of carbohydrate microarrays. This screen detects loss of activity by probing an array of enzyme/substrate mixtures with an antibody that is specific for highly esterified pectin (14). Although it might be possible to employ such an assay to screen for thermostabilizing substitutions, the assay we employed is simpler and obviates the need for the monoclonal antibody.

We identified five different amino acid residues at which thermostabilizing substitutions were found. In general, we found that the stabilizing properties of each substitution were broadly additive with the exception of the Y158H substitution. As we combined the various substitutions identified by our first-generation screen, we performed site-saturation mutagenesis to search for other substitutions at the same position that would contribute to thermostability. In no cases (T273, V251, and Y158) did we identify a new substitution that conferred greater thermostability and maintained catalytic activity than the substitution initially identified. This is somewhat surprising to us in that with the relatively low rate of introduction of mutations by error-prone PCR, the mutagenized library results overwhelmingly in single base changes in codons. Therefore, some substitutions are not attainable by this method. Nonetheless, site saturation, which does explore all of the substitutions possible at a given position, did not yield any better substitutions.

At position 273, in addition to finding the original arginine substitution, we found a substitution to lysine that conferred roughly the same added thermostability. This strongly suggests that a positively charged residue is specifically required for thermostabilization at this position in the protein. In fact, aspartate at position 282 is close enough in the three-dimensional structure of the protein to position 273 (backbone alpha carbon atoms within 5.6 Å of each other) to be the partner for an ion pair between lysine or arginine at this position. Consistent with this proposal, the molecular modeling program SCRWL ( predicts the formation of an R273/D282 salt-bridge with the T273R substitution. At position 251, we also recapitulated the first-generation substitution to alanine and found that substitution to cysteine conferred roughly the same added stability. This suggests that a small side chain is best at this position.

We found no substitutions at position 158 that conferred detectably greater stability. This is understandable, given that when the Y158H substitution was specifically added to the JL25 background, it resulted in an 0.5°C increase in thermostability at most, and our screen was for clones with thermostability gains of at least 2°C from the previous generation. Here, we found an apparent case of background-dependent stabilization because, when added to the CC4 background, the Y158H substitution confers a gain in thermostability of at least 2°C. Y158, along with Y181, F202, and W269, contributes to an aromatic region surrounding the active site (8, 10). Interestingly, in their screening of mutants, Øbro and colleagues explored substitutions at position 158 of PMEA. They found that only substitution to cysteine (and possibly aspartic acid) resulted in loss of function. Consistent with our findings, substitution to histidine did not result in a loss of function (14).

Many mechanisms have been proposed to account for thermal stabilization of proteins, and some of our substitutions are consistent with some of these proposals (17, 22, 25). The T273R and T273K substitutions are consistent with the formation of an additional ion pair on the surface of the protein. These substitutions would also limit the conformational flexibility of the prominent extension on the protein's surface. The S228F substitution, occurring at the transition of a strand to a loop, is consistent with increased stability due to improved hydrophobic packing. On the other hand, the R253K substitution would appear to be a very subtle change that can't be accounted for by any simple rules to predict thermostabilization, and yet it resulted in an approximately 5°C increase in Tm. The substitutions at residue 251 (V to A or C) also do not readily conform to any general observations on the types of substitutions that lead to stabilization.

Interestingly, the positions at which we found thermostabilizing substitutions are generally close to the substrate-binding surfaces, as identified by Fries and colleagues (8), and in the case of S228, have been identified by mutation as participating directly in substrate binding. Perhaps not surprisingly then, these residues are also among the more conserved among the alignment of PME homologs (ConSurf software, version 3.0 []). However, given the strong conservation and presumably selection for function, one might not have expected to find the stabilizing mutations at these same positions. In the cases of positions 251, 253, and 273, none of the five different thermostabilizing substitutions we found are found in any of the homologous PMEs in the ConSurf alignment. Thus, directed evolution by family shuffling (6) would likely not have found these same substitutions.

The kinetics of thermal inactivation (Fig. (Fig.5B)5B) indicate that the JL25 variant has a half-life of 13.9 min at 65°C in 0.02 M succinate (pH 6.0) buffer and that there is no diminution in activity when incubated at 62°C under these conditions. Furthermore, we have found that the JL25 variant remains active on purified citrus pectin or pectin in sugar beet pulp at 60°C for at least 24 to 48 h (data not shown). Only when the temperature exceeded 65°C did we find diminished activity on pectin substrates (Fig. (Fig.5A).5A). It is likely that the different results are due to the presence of substrate in the experiment shown in Fig. Fig.5A,5A, while it was not in the thermal inactivation kinetics shown in Fig. Fig.5B.5B. Pitkänen and colleagues noted the same effect (16). They reported a 50°C optimal temperature for E. chrysanthemi PMEA activity, but in the absence of substrate, a PMEA stability of only up to 40°C.

The thermostabilizing substitutions did not appreciably diminish the catalytic constant of JL25, 655 s−1, compared to that of the wild-type enzyme, 688 s−1. However, the Michaelis constant of JL25, 0.073%, is greater than that of the parental enzyme, 0.044%. Fortunately, for application to biomass degradation where substrate concentrations will be very high, modest increases in Km are not problematic. The ability of JL25 to work at 60 to 65°C more than compensates for any decreased catalytic activity, as its specific activity at its optimal temperature is greater than that of the wild-type enzyme at its optimal temperature. Although the Km we calculated for the wild-type enzyme is consistent with that reported by others, the calculated turnover number is greater. This could be the result of slight differences in reaction conditions or as a result of the different assay systems. We measured the rate of methanol production to assay our PME activities, while Fries et al. used a pH stat to assay the release of protons during the PME reaction (8).

Sugar beet pulp is the residue following industrial sucrose extraction from Beta vulgaris roots. Sugar beet processing involves slicing the roots into thin strips and extracting the sucrose by countercurrent diffusion with hot water. The remaining material is pressed to remove excess sucrose and water, leaving behind a 60°C, 75%-water-content pulp. This pulp is well-suited for enzymatic saccharification of the remaining polysaccharides (~24% pectin, ~32% hemicellulose, ~20% cellulose by dry weight). The resultant sugars could be used for fermentation to biofuel or for chemical conversion to other value-added products. The results shown in Table Table11 indicate that the thermostabilized JL25 PME has utility in facilitating the saccharification of the beet pulp pectin by a commercial pectinase, Pectinex Ultra SPL. Pectinex is a mixture of enzymes derived from culture filtrates of Aspergillus aculeatus (Novozymes). High-performance liquid chromatography analysis of the hydrolysis mixture suggests that without added PME, Pectinex alone produced galacturonic acid very slowly. We also found that the addition of JL25 PME substantially enhanced the rate of production of arabinose relative to Pectinex alone (data not shown). The results of sugar beet pulp hydrolysis in Table Table11 were obtained by hydrolysis at 55°C. The instability of the pectate lyase in Pectinex at temperatures above 55°C prevented us from digesting the pulp at 60° to 65°C where we would expect the JL25 PME variant to be most effective (Fig. (Fig.5A).5A). Engineering pectate lyases and polygalacturonate hydrolases to greater thermostability and activity in sugar beet pulp will further facilitate the saccharification of sugar beet pulp.


This work was supported by Small Business Innovation Research grant IIP-0638102 from the National Science Foundation and by the Maryland Technology Transfer Fund.


[down-pointing small open triangle]Published ahead of print on 9 October 2009.


1. Abbott, D. W., and A. B. Boraston. 2008. Structural biology of pectin degradation by Enterobacteriaceae. Microbiol. Mol. Biol. Rev. 72:301-316. [PMC free article] [PubMed]
2. Beaulieu, C., M. Boccara, and F. Van Gijsegem. 1993. Pathogenic behavior of pectinase-defective Erwinia chrysanthemi mutants on different plants. Mol. Plant-Microbe Interact. 6:197-202.
3. Boccara, M., and V. Chatain. 1989. Regulation and role in pathogenicity of Erwinia chrysanthemi 3937 pectin methylesterase. J. Bacteriol. 171:4085-4087. [PMC free article] [PubMed]
4. Bradford, M. M. 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. [PubMed]
5. Christgau, S., L. V. Kofod, T. Halkier, L. N. Andersen, M. Hockauf, K. Dörreich, H. Dalbøge, and S. Kauppinen. 1996. Pectin methyl esterase from Aspergillus aculeatus: expression cloning in yeast and characterization of the recombinant enzyme. Biochem. J. 319:705-712. [PubMed]
6. Crameri, A., S. A. Raillard, E. Bermudez, and W. P. Stemmer. 1998. DNA shuffling of a family of genes from diverse species accelerates directed evolution. Nature 391:288-291. [PubMed]
7. Downie, B., L. M. A. Dirk, K. A. Hadfield, T. A. Wilkins, A. B. Bennett, and K. J. Bradford. 1998. A gel diffusion assay for quantification of pectin methylesterase activity. Anal. Biochem. 264:149-157. [PubMed]
8. Fries, M., J. Ihrig, K. Brocklehurst, V. E. Shevchik, and R. W. Pickersgill. 2007. Molecular basis of the activity of the phytopathogen pectin methylesterase. EMBO J. 26:3879-3887. [PubMed]
9. Jayani, R. S., S. Saxena, and R. Gupta. 2005. Microbial pectinolytic enzymes: a review. Process Biochem. 40:2931-2944.
10. Jenkins, J., O. Mayans, D. Smith, K. Worboys, and R. W. Pickersgill. 2001. Three-dimensional structure of Erwinia chrysanthemi pectin methylesterase reveals a novel esterase active site. J. Mol. Biol. 305:951-960. [PubMed]
11. Laurent, F., A. Kotoujansky, and Y. Bertheau. 2000. Overproduction in Escherichia coli of the pectin methylesterase A from Erwinia chrysanthemi 3937: one-step purification, biochemical characterization and production of polyclonal antibodies. Can. J. Microbiol. 46:474-480. [PubMed]
12. Laurent, F., A. Kotoujansky, G. Labesse, and Y. Bertheau. 1993. Characterization and overexpression of the pem gene encoding pectin methylesterase of Erwinia chrysanthemi strain 3937. Gene 131:17-25. [PubMed]
13. Miller, G. L. 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugars. Anal. Chem. 31:426-428.
14. Øbro, J., I. Sørensen, P. Derkx, C. T. Madsen, M. Drews, M. Willer, J. D. Mikkelsen, and W. G. T. Willats. 2009. High-throughput screening of Erwinia chrysanthemi pectin methylesterase variants using carbohydrate microarrays. Proteomics 9:1861-1868. [PubMed]
15. Pelloux, J., C. Rustérucci, and E. J. Mellerowicz. 2007. New insights into pectin methylesterase structure and function. Trends Plant Sci. 12:267-277. [PubMed]
16. Pitkänen, K., R. Heikinheimo, and R. Pakkanen. 1992. Purification and characterization of Erwinia chrysanthemi B374 pectin methylesterase produced by Bacillus subtilis. Enzyme Microb. Technol. 14:832-836.
17. Razvi, A., and J. M. Scholtz. 2006. Lessons in stability from thermophilic proteins. Protein Sci. 15:1569-1578. [PubMed]
18. Ridley, B. L., M. A. O'Neill, and D. Mohnen. 2001. Pectins: structure, biosynthesis and oligogalacturonide-related signaling. Phytochemistry 57:929-967. [PubMed]
19. Schols, H. A., and A. G. J. Voragen. 2002. The chemical structure of pectins, p. 1-30. In G. B. Seymnour and J. P. Knox (ed.), Pectins and their manipulation. CRC Press, Boca Raton, FL.
20. Shevchik, V. E., G. Condemine, N. Hugouvieux-Cotte-Pattat, and J. Robert-Baudouy. 1996. Characterization of pectin methylesterase B, an outer membrane lipoprotein of Erwinia chrysanthemi 3937. Mol. Microbiol. 19:455-466. [PubMed]
21. Shevchik, V. E., and N. Hugouvieux-Cotte-Pattat. 1997. Identification of a bacterial pectin acetyl esterase in Erwinia chrysanthemi 3937. Mol. Microbiol. 24:1285-1301. [PubMed]
22. Szlágyi, A., and P. Závodszky. 2000. Structural differences between mesophilic, moderately thermophilic and extremely thermophilic protein subunits: results of a comprehensive survey. Structure 8:493-504. [PubMed]
23. Valette-Collet, O., A. Cimerman, P. Reignault, C. Levis, and M. Boccara. 2003. Disruption of Botrytis cinerea pectin methylesterase gene Bcpme1 reduces virulence on several host plants. Mol. Plant-Microbe Interact. 16:360-367. [PubMed]
24. Willats, W. G. T., J. P. Knox, and J. Dalgaard Mikkelsen. 2006. Pectin: new insights into an old polymer are starting to gel. Trends Food Sci. Technol. 17:97-104.
25. Wintrode, P. L., and F. H. Arnold. 2000. Temperature adaptation of enzymes: lessons from laboratory evolution. Adv. Protein Chem. 55:161-225. [PubMed]
26. Wojciechowski, C. L., and R. Fall. 1996. A continuous fluorometric assay for pectin methylesterase. Anal. Biochem. 237:103-108. [PubMed]

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)