The Structural Basis for Exopolygalacturonase Activity in a Family 28 Glycoside Hydrolase

J. Mol. Biol. (2007) 368, 1215–1222

doi:10.1016/j.jmb.2007.02.083

COMMUNICATION

The Structural Basis for Exopolygalacturonase Activity in a Family 28 Glycoside Hydrolase D. Wade Abbott and Alisdair B. Boraston⁎ Biochemistry and Microbiology, University of Victoria, PO Box 3055 STN CSC, Victoria BC, Canada V8W 3P6

Family 28 glycoside hydrolases (polygalacturonases) are found in organisms across the plant, fungal and bacterial kingdoms, where they are central to diverse biological functions such as fruit ripening, biomass recycling and plant pathogenesis. The structures of several polygalacturonases have been reported; however, all of these enzymes utilize an endo-mode of digestion, which generates a spectrum of oligosaccharide products with varying degrees of polymerization. The structure of a complementary exo-acting polygalacturonase and an accompanying explanation of the molecular determinants for its specialized activity have been noticeably lacking. We present the structure of an exopolygalacturonase from Yersinia enterocolitica, YeGH28 in a native form (solved to 2.19 Å resolution) and a digalacturonic acid product complex (solved to 2.10 Å resolution). The activity of YeGH28 is due to inserted stretches of amino acid residues that transform the active site from the open-ended channel observed in the endopolygalacturonases to a closed pocket that restricts the enzyme to the exclusive attack of the non-reducing end of oligogalacturonide substrates. In addition, YeGH28 possesses a fused FN3 domain with unknown function, the first such structure described in pectin active enzymes. © 2007 Published by Elsevier Ltd.

*Corresponding author

Keywords: pectin degradation; glycoside hydrolase; exopolygalacturonase; Yersinia enterocolitica; X-ray crystallography

Family 28 polygalacturonases (GH28) are glycosidases that hydrolyze the homogalacturonan and rhamnogalacturonan components of pectin. Both of these reactions occur via a single displacement inverting mechanism, which results in fully saturated products with an altered stereochemistry around the anomeric carbon. These enzymes are mechanistically different from the other major class of pectinase: the pectate and pectin lyases, which cleave glycosidic linkages by β-elimination, resulting in products with a 4,5-unsaturation at their nonreducing end. Several three-dimensional structures of secreted endo-active GH28s have been presented, including examples from both bacterial1 and fungal species.2–7 The most noticeable structural feature of these enzymes is their signature right-handed β-helix fold that was first identified in the family 1 pectate lyase, PelC.8,9 This fold is now understood to be a E-mail address of the corresponding author: [email protected] 0022-2836/$ - see front matter © 2007 Published by Elsevier Ltd.

scaffold highly utilized by pectinolytic enzymes in general.10 Polygalacturonases typically display ten complete turns of β-structure, which is composed of four parallel sheets extending along the longitudinal axis. This latter observation distinguishes them from the polysaccharide lyases, which generally contain a three-sheet topology. The active site of polygalacturonases is a surface channel that is open at both ends enabling the enzyme to attack internal stretches of sugars within a polysaccharide. By definition, these enzymes produce oligogalacturonic acid products with various degrees of polymerization. Although the activity of endopolygalacturonases is well understood, the structural features that determine exopolygalacturonase activity for related enzymes have remained unresolved. 11 Exopolygalacturonases are unable to macerate plant tissue, and are believed to play an important role in the later stages of pectin processing. In Erwinia sp. and the Gram-negative enteric pathogen Yerisinia enterocolitica, exo-acting GH28s are present within the periplasmic space and function to convert longer oligogalacturonide substrates

1216 exclusively into digalacturonic acid (diGalUA).12–14 The Erwinia ortholog of YeGH28 is preferentially active on substrates with a 4,5-unsaturation (such as those generated by pectate lyases), ultimately resulting in unsaturated digalacturonides for intracellular transport.12 We have recently observed that the dedicated specificity determinant for the TogMNAB oligogalacturonide ABC transporter, TogB, interacts preferentially with digalUA ligands, showing a higher affinity for the unsaturated form. 42 By generating the optimal ligand for intracellular transport, periplasmic GH28s have an important role in pectin catabolism by operating as a functional bridge between extracellular and intracellular pectin catabolism. Here, we present the novel structure of the GH28 from Y. enterocolitica (YeGH28) in a native form and in complex with a diGalUA product. This structural information illuminates for the first time the molecular basis of its exclusive exopolygalacturonase activity. Cloning and purification of YeGH28 The mature sequence of YeGH28 (with the Nterminal signal peptide removed) was PCR amplified from genomic Y. enterocolitica DNA (ATCC 9610D) using the primer sequences: 5′-ATATGCTAGCGCAAAATCAAGTAGTCTCGATGCTCCT-3 5′-ATATAAGCTTTTAAGGCGCGACAGGTTTACCATCA-3 The PCR fragments were restricted by 5′ NheI and 3′ XhoI digestion, and cloned into pET28a (Novagen, La Jolla, CA; cat. no. 69864-3) as described.15 The YeGH28 construct was then transformed into BL21 DE3 star cells and expression induced by 0.2 mM IPTG within a 3.0 l culture at an absorbance at 600 nm of 0.8. All attempts at producing the protein in a soluble form were unsuccessful, so a method of refolding was pursued. Induced cells were grown to stationary phase and harvested by centrifugation, lysed by French press, and the cell debris pelleted by centrifugation. The pellet was washed twice with sterile distilled water and then three times with 20 mM Tris–HCl (pH 8.0), 0.5% (v/v) Triton-X-100, 500 mM NaCl. Finally, the detergent was removed by washing twice with 20 mM Tris–HCl (pH 8.0), 500 mM NaCl. The insoluble protein pellet was then dissolved in 30 ml of 20 mM Tris–HCl (pH 8.0), 6.0 M urea, 500 mM NaCl, and stirred vigorously at room temperature for 1 h. This solution was clarified by centrifugation at 10,000g for 45 min and loaded onto an IMAC column equilibrated in 20 mM Tris–HCl (pH 8.0), 6.0 M urea, 500 mM NaCl. Protein was eluted by a 5 mM–500 mM imidazole gradient in the same buffer conditions. YeGH28 was analyzed for purity by SDS-PAGE, and samples containing appreciable amounts of protein were pooled and concentrated to ∼5 mg/ml. Urea was removed and refolding enabled by exhaustive dialysis against

Structure of an Exo-acting GH28

20 mM Tris–HCl (pH 8.0) at 4 °C. During refolding, an estimated 95% of the sample was lost to precipitation; however, the remaining, properly folded protein was stable and crystallized in several conditions. Crystallization, data collection, and structure solution YeGH28 at 10 mg/ml was crystallized in space group P21 by the hanging-drop, vapor-diffusion method at 18 °C in 20% (w/v) PEG 4000, 0.1 M nickel sulfate, 0.1 M sodium acetate, pH 4.2. YeGH28 crystals were frozen at 100 K directly in the cryostream after a brief soak in mother liquor supplemented with 15% (v/v) ethylene glycol. A SAD data set at the Ni2+ absorption edge, determined by a fluorescence scan, was collected at the National Synchrotron Light Source (NSLS, Brookhaven National Laboratories). Diffraction data were processed using Crystal Clear/d*trek.16 ShelxD17 found nine heavy atom sites, which were used for SAD phasing with SHARP.18 Refinement of the nine sites using data to 2.19 Å yielded an anomalous phasing power of 0.8 and figures of merit of 0.24 and 0.6 for acentric and centric reflections, respectively. Noncrystallographic symmetry operators defining the relationship between the two YeGH28 molecules in the asymmetric unit were determined from the heavy-atom sites using profess.19 Non-crystallographic symmetry averaging and solvent flattening with DM using a solvent content of 53% resulted in a figure of merit of 0.82 and readily interpretable maps.20 ARP/wARP was able to correctly build a virtually complete model with docked side-chains for both monomers in the asymmetric unit.21 The model was then completed manually using COOT,22 and refined with REFMAC.23 The diGalUA complex was generated by soaking YeGH28 crystals for 2 h at 16 °C with 15 mM trigalacturonic acid dissolved in mother liquor. The complex was solved by refining the YeGH28 native model against the diGalUA dataset. Final coordinates were validated using PROCHECK24 and SFCHECK25 in CCP4.26 All data collection, processing, and final model statistics are given in Table 1. Water molecules were added using the ARP/wARP option within REFMAC and inspected visually before deposition.21 A randomly chosen 5% of the reflections were flagged as free and used to monitor the refinement progress.27 Overall fold of YeGH28 YeGH28 crystallized as a dimer (chains A and B in the asymmetric unit); although gel-filtration chromatography suggested that it is a monomer in solution. Both monomers in the asymmetric unit contained loops with poorly defined electron density. The loop comprising residues D503–N508 in chain B proved too disordered to be modeled, while this loop could be built in chain A, allowing for a complete model of chain A. The overall structure of YeGH28 is presented in Figure 1(a). The protein adopts a conventional right-handed parallel β-helix

Structure of an Exo-acting GH28

1217

Table 1. Crystallography statistics YeGH28 A. Data collection Space group Cell parameters a (Å) b (Å) c (Å) α (deg.) β (deg.) γ (deg.) Resolution (Å) Rmerge I/σI Completeness (%) Redundancy B. Refinement Resolution (Å) No. reflections Rwork Rfree No. atoms Protein ADA Ni SO4 Acetate Water B-factors Protein (Å2) ADA (Å2) Ni (Å2) SO4 (Å2) Acetate (Å2) Water (Å2) rmsd from ideal Bond lengths (Å) Bond angles (deg.) Ramachandran plot regions Allowed (%) Generously allowed (%) Disallowed (%)

Native

DiGalUA complex

P21

P21

89.95 78.76 98.19 90.00 103.60 90.00 20.00–2.19 (2.27–2.19) 0.095 (0.286) 11.9 (5.3) 99.4 (94.5) 7.3 (8.8)

91.01 79.65 98.49 90 103.90 90.00 19.96–2.10 (2.17–2.10) 0.11 (0.48) 4.5 (1.2) 99.3 (99.3) 2.2 (2.2)

20.0–2.19 64,794 0.18 0.23 9409 8838 – 10 95 4 462

19.72–2.10 75,070 0.23 0.28 9468 8826 1 10 21 4 487

24.4 – 72.9 52.9 20.0 28.0

34.2 47.7 56.7 73.7 43.1 32.9

0.019 1.82

0.014 1.58

98.2 1.2 0.6

97.8 1.4 0.8

actually known about their function within this context. Traditionally, FN3s are involved in a variety of molecular recognition process such as cell adhesion, cell surface hormone and cytokine receptors, and chaperonins by facilitating protein–protein interactions.28 Dali structure alignments29 revealed that the YeGH28 FN3 domain has the most structural similarity with the human fibronectin-binding protein,30 overlapping with an rmsd of 1.38 Å (for 82 aligned Cα) (Figure 1(b)). Between the two motifs, the characteristic seven-stranded β-sandwich is highly conserved; however, the YeGH28 FN3 displays a helix-loop-helix motif between β-strands 4 and 5. This protrusion appears to assist in anchoring the FN3 domain onto the enzyme core, as there are many hydrophobic side-chains in contact between the surfaces of these two structures. Although the biological role of the FN3 module in oligogalacturonide degradation remains to be determined, the fact that it is fused to the opposite side of the enzyme than the active site suggests that (1) it is not involved directly in catalysis and (2) interactions with other molecules mediated by the FN3 module would not occlude accessibility to the YeGH28 catalytic machinery. As YeGH28 is a monomer in solution (data not shown), the FN3

Values within parentheses are for the highest resolution shell.

topology, containing ten complete turns that comprise four discernable β-sheets. The β-helix is capped with an N-terminal α-helix and an extended Cterminal tail that lacks any defined secondary structure. YeGH28 is a 570 amino acid residue protein, which is much larger than the majority of reported pectinases, including the endo-acting polygalacturonases; an observation that is attributable to the fibronectin-type III (FN3) domain grafted to its N terminus (Figure 1(b)). Apart from its two distinct structural features, the FN3 domain and active site loops (described in more detail below), YeGH28 displays significant sequence and structural similarity with other GH28s (Figure 2). Importantly, the eight previously defined residues involved in substrate recognition and catalysis are stringently conserved within this enzyme family.1,4,5 The FN3 domain FN3 domains are quite common in bacterial carbohydrate active proteins; however, very little is

Figure 1. The three-dimensional structure of the exopolygalacturonase YeGH28. (a) Chain A of YeGH28 is displayed in ribbon format and colored in a rainbow scheme. (b) Superimposition of the YeGH28 FN3 domain (blue) with the third FN3 domain of fibronectin (beige, PDB ID 1FNF).

1218

Structure of an Exo-acting GH28

Figure 2. Comparison of family 28 glycoside hydrolases. Secondary structure alignment of YeGH28 with the closely related endopolygalacturonases from S. purpureum and E. carotovora. Alignments were generated by superimposing enzyme structures and mapping the primary sequences. These elements were then aligned by hand. The loop insertions (red) and accessory structures (green) are indicated below the sequences. The loops highlight the regions within the active site that confer exo-activity; the α-helix is an extended region with YeGH28 that adopts a helix-turn-helix-turn-helix topology, which supports loop 1; the β-sheet is a two-stranded antiparallel β-structure with unknown function; and the C-terminal cap refers to an extended region at the C terminus of the molecule with no defined secondary structure. Conserved catalytic residues are indicated with a filled triangle.

domain would not appear to be involved in the formation of homomultimers. Recently, an interesting study characterized the cellulose-binding properties of the Clostridium thermocellum cellobiohydrolase FN3. 31 The authors identified a novel role for this domain in cellulose structure modification and enzyme efficiency, suggesting a possible alternative role for the FN3 of YeGH28 in carbohydrate recognition. In the YeGH28 FN3, however, there is no detectable surfaceaccessible aromatic amino acid side chain, which is

classically associated with carbohydrate-binding site formation. Additionally, there was no detectable electron density for an oligogalacturonide ligand following crystal soaking experiments. Even in the light of these data, a potential role for the YeGH28 FN3 in carbohydrate recognition should not be ruled out for two reasons: (1) structural studies suggest that two different galacturonic acid-binding proteins utilize electrostatic interactions during complex formation42 and, therefore, ligand binding may be facilitated by polar amino acids; and (2) the

Structure of an Exo-acting GH28

orientation of the protein within the unit cell may restrict access of a ligand to a potential binding site. Conserved catalytic machinery In order to characterize the architecture of the active site in the exopolygalacturonase, we soaked YeGH28 crystals with trigalacturonic acid ligand. Excellent electron density was observed for a complete digalacturonic acid product molecule in the active site of chain A, confirming that the enzyme is functional and folded properly. This information has allowed us to compare the details of the YeGH28 catalytic site with those of its endo-acting counterparts. The active sites of several GH28s have been mapped in Erwinia carotovora,1 Aspergillus niger,4 and Stereum purpureum.5 A constellation of three catalytic aspartate residues: D202, D223 and D224 (E. carotovora) and D153, D17 3 and D174 (S. purpureum) has been characterized by site-directed mutagenesis,1,4 and product complex studies,5 respectively. These residues are involved in protonation and deprotonation events across the reaction coordinate, and mutation of any of them ablates catalytic activity.4 Polygalacturonases utilize a distinctive syn inverting mechanism to depolymerize polygalacturonic acid substrates, where the catalytic aspartate residues are

1219 clustered within ∼ 5 Å of each other and approach the substrate from the same side (Figure 3(a)). Structural alignments of YeGH28 with the homologous polygalacturonases from E. carotovora and S. purpureum suggest that the catalytic aspartate residues are functionally conserved between exo and endo-acting polygalacturonases (Figure 3(a)). Indeed, D381, D402, and D403 from YeGH28, which are all within 5 Å of one another, align very well with the analogous residues in the other structures. The digalacturonic product complex is orientated with its reducing end residue presenting its α-face towards the catalytic cluster, which is consistent with the S. purpureum ternary complex.5 The scissile glycosidic oxygen is within 3.3 Å of the catalytic acid D402 (Figure 3(a) and (b)). Although there is no water molecule interacting with D381 and D403 in the product complex, overlapping the native structure with the complex introduces a water molecule into the active site that is in hydrogen bond contact with D403 and has satisfactory geometry and position to attack the anomeric carbon. Analysis of the interactions between diGalUA and YeGH28 reveals that there are several basic amino acids involved in stabilizing the residue in the − 2 subsite (Figure 3(b)), which also function to enclose the non-reducing end of the active site and position the substrate properly for hydrolysis (Figure 3(c)).

Figure 3. Active site architecture of YeGH28. (a) Divergent wall-eyed view of overlapped polygalacturonases from Y. enterocolitica (blue), E. carotovora (grey, PDB ID 1BHE), and S. purpureum (yellow, PDB ID 1KCD). The electron density around the digalacturonic acid is represented as maximum-likelihood 23/σa41 weighted 2Fobs–Fcalc map contoured at 0.8 σ (0.21 e/Å3). (b) A representation of the YeGH28 −1 and −2 subsite interactions with diGalUA. (c) Solvent-accessible surface representation of the active site in complex with diGalUA viewed from the reducing end of the sugar.

1220

Structure of an Exo-acting GH28

O-2 is in hydrogen bond contact with R440 and N406, an amino acid that concomitantly interacts with O-3. The C-5 uronate group, a definitive substituent in galacturonic acid, forms an H bond with H335 and a salt-bridge with R240. Interestingly, electrostatic interactions are a repeating theme in galacturonic acid recognition within periplasmic proteins and we have recently identified examples from the pectate lyase YePL2A and ABC transporter specificity determinant TogB.42 It is important to note that there is no detectable interaction with the axial O-4 of diGalUA, which also is a structural determinant for galacto-configured sugars. This observation is in agreement with a previous report that the Erwinia sp. GH28 ortholog has a higher specificity for substrates with a 4,5-unsaturation. Modeling an unsaturated diGalUA molecule into the electron density for the diGalUA complex, however, did not reveal noticeable differences in potential binding profiles between the two sugars. There are surprisingly few interactions within the − 1 subsite. The diGalUA forms several H bonds within a solvation network and the O-1 group interacts with a nickel ion that is an artefact of the crystallization buffer (not shown). Further analysis of active site architecture will greatly benefit from mutagenesis and substrate co-crystallization. The structural basis of exopolygalacturonase activity The active site of endopolygalacturonases is best described as a channel (Figure 4(b)). The open ends of the active site enable the enzyme to cleave polysaccharide chains internally as the reducing and non-reducing ends of the carbohydrate can extend outward towards the bulk solvent in either direction. Such reactions result in the generation of carbohydrate products with various degrees of polymerization. In many cases, enzyme processivity is enabled also by the enzyme being restricted in its ability to release its substrate. Previously, YeGH28 was characterized as an exclusive diGalUA-producing exopolygalacturonase by product profiling.13 This observation is in agreement with the activity of the highly similar (61% identical) enzyme PehX from Erwinia chrysanthemi.13,32 However, the structural determinants of substrate selectivity and positioning are unresolved. The structure solution of YeGH28 in complex with a diGalUA product presented here allows us to elucidate how this process occurs. The secondary structure alignment of YeGH28, EcGH28, and SpGH28 highlights the insertion of several novel elements, including a substantial αhelix and β-sheet motif important for stabilizing active site loops and the FN3 domain, respectively, and four key loops that seal off the non-reducing end of the active site (Figures 2 and 4). Interestingly, loop 1 (residues 225–242) is actually conserved between YeGH28 and EcGH28; however, within the endopolygalacturonase structure this loop is orientated away from the active site and contributes to a wall of the channel. Loops 2–4, on the other

Figure 4. Structural basis of exopolygalacturonase activity. (a) Superimposition of polygalacturonases from Y. enterocolitica (blue) and E. carotovora (grey). The active sites are positioned at the top of the structures. Novel secondary structures found in YeGH28 are highlighted as in Figure 2: loops (1–4), red; accessory structures (FN3, αhelix, β-sheet, and C-terminal cap), green. (b) Solventaccessible surface model of EcGH28 displaying the tunnellike active site. The catalytic aspartate residues D202, D223, and D224 are shaded in red. (c) Solvent-accessible surface model of YeGH28, displaying the pocket-like active site. The three putative catalytic residues D381, D402 and D403 are shaded in red and the novel loop insertions, which seal off the active site, are highlighted in blue.

Structure of an Exo-acting GH28

hand, are unique to the YeGH28 exopolygalacturonase and seal off the end of the active site where the non-reducing end of a sugar is accommodated (Figure 4(c)). B-factor analysis of YeGH28 in its native and complex forms reveals that the protein has a high level of flexibility within loop 4, consistent with the crystal producing very little electron density in chain B of the native form and in both monomers of the product complex (not shown). The other three loop insertions have a lower degree of flexibility within the crystal structure, suggesting that the pocket-like active site within YeGH28 is in a stabilized form. The solvent-accessible surface representation of YeGH28 highlights its pocket-like active site structure, which is tailored to attack oligogalacturonides at their non-reducing ends (Figures 3(d) and 4(d)). This topography sterically prevents access of the enzyme to the internal residues of polygalacturonic acid. Closer analysis of the active site reveals the structural determinants for the exclusive production of disaccharide products. Two well-defined subsites accommodate oligogalacturonide substrates at positions − 1 and − 2, and orient the scissile bond within the proximity of the catalytic machinery (Figure 3). Substrate specificity is likely determined also by subsites beyond the catalytic site, as a channel lined with basic amino acids extends outward and across the surface of the enzyme (Figure 4(d)). In this manner, YeGH28 may contain inherent processivity as oligogalacturonide substrates are fed into the active site pocket and degraded continuously from their non-reducing ends. This possibility is agreement with the demonstrated activity of YeGH28 on polygalacturonic acid.13 Conclusions The structure of YeGH28 provides new insight into the activity of the periplasmic exopolygalacturonase found in many pectinolytic bacteria. Amino acid insertions (loops 2–4) form a wall at the nonreducing end of the active site and block access of the enzyme to the internal residues of a polysaccharide substrate. This process of active site transformation, from endo into exo activity by loop insertions, appears to be a widely occurring phenomenon that has been reported for various glycoside hydrolase families with very different specificities.33–40 YeGH28 activity (and by association other GH28 exogalacturonase homologs) is an excellent fit with its periplasmic localization and functional position within the pectin utilization pathway of pectinolytic Enterobacteriaciae. Directed exo-digestion of oligogalacturonides,13 with specificity for 4,5-unsaturated products,12 generates saturated and unsaturated diGalUA, the preferential ligands for TogMNAB-mediated periplasmic transport. In this fashion, YeGH28 bridges the polymeric breakdown activities of upstream enzymes, and the intracellular accumulation of galacturonic acid metabolites.

1221 Protein Data Bank accession codes The atomic coordinates and structure factors have been deposited with the PDB codes of 2UVE and 2UVF for the uncomplexed and complexed structures of YeGH28, respectively.

References 1. Pickersgill, R., Smith, D., Worboys, K. & Jenkins, J. (1998). Crystal structure of polygalacturonase from Erwinia carotovora ssp. carotovora. J. Biol. Chem. 273, 24660–24664. 2. Cho, S. W., Lee, S. & Shin, W. (2001). The X-ray structure of Aspergillus aculeatus polygalacturonase and a modeled structure of the polygalacturonaseoctagalacturonate complex. J. Mol. Biol. 311, 863–878. 3. van Pouderoyen, G., Snijder, H. J., Benen, J. A. & Dijkstra, B. W. (2003). Structural insights into the processivity of endopolygalacturonase I from Aspergillus niger. FEBS Letters, 554, 462–466. 4. van Santen, Y., Benen, J. A., Schroter, K. H., Kalk, K. H., Armand, S., Visser, J. & Dijkstra, B. W. (1999). 1.68-Å crystal structure of endopolygalacturonase II from Aspergillus niger and identification of active site residues by site-directed mutagenesis. J. Biol. Chem. 274, 30474–30480. 5. Shimizu, T., Nakatsu, T., Miyairi, K., Okuno, T. & Kato, H. (2002). Active-site architecture of endopolygalacturonase I from Stereum purpureum revealed by crystal structures in native and ligand-bound forms at atomic resolution. Biochemistry, 41, 6651–6659. 6. Federici, L., Caprari, C., Mattei, B., Savino, C., Di Matteo, A., De Lorenzo, G. et al. (2001). Structural requirements of endopolygalacturonase for the interaction with PGIP (polygalacturonase-inhibiting protein). Proc. Natl Acad. Sci. USA, 98, 13425–13430. 7. Petersen, T. N., Kauppinen, S. & Larsen, S. (1997). The crystal structure of rhamnogalacturonase A from Aspergillus aculeatus: a right-handed parallel beta helix. Structure, 5, 533–544. 8. Yoder, M. D., Lietzke, S. E. & Jurnak, F. (1993). Unusual structural features in the parallel beta-helix in pectate lyases. Structure, 1, 241–251. 9. Yoder, M. D., Keen, N. T. & Jurnak, F. (1993). New domain motif: the structure of pectate lyase C, a secreted plant virulence factor. Science, 260, 1503–1507. 10. Jenkins, J. & Pickersgill, R. (2001). The architecture of parallel beta-helices and related folds. Prog. Biophys. Mol. Biol. 77, 111–115. 11. Jenkins, J., Mayans, O. & Pickersgill, R. (1998). Structure and evolution of parallel beta-helix proteins. J. Struct. Biol. 122, 236–246. 12. Hugouvieux-Cotte-Pattat, N., Condemine, G., Nasser, W. & Reverchon, S. (1996). Regulation of pectinolysis in Erwinia chrysanthemi. Annu. Rev. Microbiol. 50, 213–257. 13. Liao, C. H., Revear, L., Hotchkiss, A. & Savary, B. (1999). Genetic and biochemical characterization of an exopolygalacturonase and a pectate lyase from Yersinia enterocolitica. Can. J. Microbiol. 45, 396–403. 14. Rodionov, D. A., Gelfand, M. S. & Hugouvieux-CottePattat, N. (2004). Comparative genomics of the KdgR regulon in Erwinia chrysanthemi 3937 and other gamma-proteobacteria. Microbiology, 150, 3571–3590. 15. Boraston, A. B., Warren, R. A. & Kilburn, D. G. (2001).

Structure of an Exo-acting GH28

1222

16. 17. 18.

19. 20. 21. 22. 23.

24.

25.

26. 27. 28. 29. 30.

Glycosylation by Pichia pastoris decreases the affinity of a family 2a carbohydrate-binding module from Cellulomonas fimi: a functional and mutational analysis. Biochem. J. 358, 423–430. Pflugrath, J. W. (1999). The finer things in X-ray diffraction data collection. Acta Crystallogr. D Biol. Crystallogr. 55, 1718–1725. Schneider, T. R. & Sheldrick, G. M. (2002). Substructure solution with SHELXD. Acta Crystallog. sect. D, 58, 1772–1779. Evans, G. & Bricogne, G. (2002). Triiodide derivatization and combinatorial counter-ion replacement: two methods for enhancing phasing signal using laboratory Cu Kalpha X-ray equipment. Acta Crystallog. sect. D, 58, 976–991. Collaborative Computational Project, Number 4 (1994). The CCP4 suite: programs for protein crystallography. Acta Crystallog. sect. D, 50, 760–763. Cowtan, K. (1999). Error estimation and bias correction in phase-improvement calculations. Acta Crystallog. sect. D, 55, 1555–1567. Perrakis, A., Harkiolaki, M., Wilson, K. S. & Lamzin, V. S. (2001). ARP/wARP and molecular replacement. Acta Crystallog. sect. D, 57, 1445–1450. Emsley, P. & Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallog. sect. D, 60, 2126–2132. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. (1997). Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallog. sect. D, 53, 240–255. Laskowski, R. A., MacArthur, M. W., Moss, D. S. & Thorton, J. M. (1993). PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallog. 26, 283–291. Vaguine, A. A., Richelle, J. & Wodak, S. J. (1999). SFCHECK: a unified set of procedures for evaluating the quality of macromolecular structure-factor data and their agreement with the atomic model. Acta Crystallog. sect. D, 55, 191–205. Potterton, E., Briggs, P., Turkenburg, M. & Dodson, E. (2003). A graphical user interface to the CCP4 program suite. Acta Crystallog. sect. D, 59, 1131–1137. Brunger, A. T. (1992). Free R-value - a novel statistical quantity for assessing the accuracy of crystal-structures. Nature, 355, 472–475. Skerra, A. (2000). Engineered protein scaffolds for molecular recognition. J. Mol. Recogn. 13, 167–187. Holm, L. & Sander, C. (1993). Protein structure comparison by alignment of distance matrices. J. Mol. Biol. 233, 123–138. Leahy, D. J., Aukhil, I. & Erickson, H. P. (1996). 2.0 Å crystal structure of a four-domain segment of human

31.

32.

33. 34.

35.

36.

37.

38.

39.

40.

41. 42.

fibronectin encompassing the RGD loop and synergy region. Cell, 84, 155–164. Kataeva, I. A., Seidel, R. D., 3rd, Shah, A., West, L. T., Li, X. L. & Ljungdahl, L. G. (2002). The fibronectin type 3-like repeat from the Clostridium thermocellum cellobiohydrolase CbhA promotes hydrolysis of cellulose by modifying its surface. Appl. Environ. Microbiol. 68, 4292–4300. He, S. Y. & Collmer, A. (1990). Molecular cloning, nucleotide sequence, and marker exchange mutagenesis of the exo-poly-alpha-D-galacturonosidaseencoding pehX gene of Erwinia chrysanthemi EC16. J. Bacteriol. 172, 4988–4995. Davies, G. & Henrissat, B. (1995). Structures and mechanisms of glycosyl hydrolases. Structure, 3, 853–859. Varrot, A., Schulein, M. & Davies, G. J. (1999). Structural changes of the active site tunnel of Humicola insolens cellobiohydrolase, Cel6A, upon oligosaccharide binding. Biochemistry, 38, 8884–8891. Varrot, A., Hastrup, S., Schulein, M. & Davies, G. J. (1999). Crystal structure of the catalytic core domain of the family 6 cellobiohydrolase II, Cel6A, from Humicola insolens, at 1.92 Å resolution. Biochem. J. 337, 297–304. Rouvinen, J., Bergfors, T., Teeri, T., Knowles, J. K. & Jones, T. A. (1990). Three-dimensional structure of cellobiohydrolase II from Trichoderma reesei. Science, 249, 380–386. Schubot, F. D., Kataeva, I. A., Chang, J., Shah, A. K., Ljungdahl, L. G., Rose, J. P. & Wang, B. C. (2004). Structural basis for the exocellulase activity of the cellobiohydrolase CbhA from Clostridium thermocellum. Biochemistry, 43, 1163–1170. Fushinobu, S., Hidaka, M., Honda, Y., Wakagi, T., Shoun, H. & Kitaoka, M. (2005). Structural basis for the specificity of the reducing end xylose-releasing exooligoxylanase from Bacillus halodurans C-125. J. Biol. Chem. 280, 17180–17186. Proctor, M. R., Taylor, E. J., Nurizzo, D., Turkenburg, J. P., Lloyd, R. M., Vardakou, M. et al. (2005). Tailored catalysts for plant cell-wall degradation: redesigning the exo/endo preference of Cellvibrio japonicus arabinanase 43A. Proc. Natl Acad. Sci. USA, 102, 2697–2702. Dias, F. M., Vincent, F., Pell, G., Prates, J. A., Centeno, M. S., Tailford, L. E. et al. (2004). Insights into the molecular determinants of substrate specificity in glycoside hydrolase family 5 revealed by the crystal structure and kinetics of Cellvibrio mixtus mannosidase 5A. J. Biol. Chem. 279, 25517–25526. Read, R. J. (1986). Improved Fourier coefficients for maps using phases from partial structures with errors. Acta Crystallog. sect. A, 42, 140–149. doi:10.1016/j.jmb.2007.03.045.

Edited by M. Guss (Received 4 December 2006; received in revised form 21 February 2007; accepted 21 February 2007) Available online 6 March 2007