View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Caltech Authors - Main research papers Acta Crystallographica Section D Biological Structure of pectate lyase A: comparison to other Crystallography isoforms ISSN 0907-4449 Leonard M. Thomas,² Chuong N. Pectate lyase A is a virulence factor secreted by the plant- Received 5 December 2001 Doan,³ Randall L. Oliver§ and pathogenic bacteria Erwinia chrysanthemi. The enzyme Accepted 2 April 2002 Marilyn D. Yoder* cleaves the glycosidic bond of pectate polymers by a calcium-dependent -elimination mechanism. The crystal PDB References: C2 PelA, structure of pectate lyase A from E. chrysanthemi EC16 has 1jta, r1jtasf; R3 PelA, 1jrg, School of Biological Sciences, University of been determined in two crystal forms, monoclinic C2 to 1.8 AÊ r1jrgsf. Missouri-Kansas City, 5007 Rockhill Road, R Ê Kansas City, MO 64110-2499, USA and rhombohedral 3 to 2.1 A. The protein structure is compared with two other pectate lyase isoforms from E. chrysanthemi EC16, pectate lyase C and pectate lyase E. ² Current address: Howard Hughes Medical Pectate lyase A is unique as it is the only acidic pectate lyase Institute, Division of Biology, California Institute and has end products that are signi®cantly more varied in of Technology, Pasadena, CA 91125, USA. ³ Current address: The Stowers Institute for length in comparison to those of the other four major pectate Medical Research, 1000 East 50th Street, lyase isozymes. Differences and similarities in polypeptide Kansas City, MO 64110, USA. trace, size and volume of the active-site groove and surface § Current address: Zymogenetics, 1201 Eastlake electrostatics are discussed. Avenue East, Seattle, WA 98102, USA. Correspondence e-mail: [email protected] 1. Introduction Pectate lyases (Pels) are virulence factors of necrotrophic pathogens that cause bacterial soft rots in plants (Collmer & Keen, 1986; Alfano & Collmer, 1996). Erwinia species are one of the most extensively characterized of the soft-rot pathogens and secrete several pectate-degrading enzymes, including pectate lyases, polygalacturonases and pectin methylesterases. Pels cleave the -1,4-galacturonsyl linkages in pectate poly- mers, which are components of both the plant cell wall and the middle lamella. The enzymes utilize a calcium-dependent -elimination mechanism. E. chrysanthemi produces ®ve major isozymes of pectate lyase, named alphabetically in order of increasing pI: PelA, PelB, PelC, PelD and PelE. The major pel genes in Erwinia are organized as two multigene families on separate chromosomal clusters. The pelA, pelE and partially deleted pelD genes of E. chrysanthemi EC16 form one cluster, presumed to be created by a dupli- cation event. The second gene cluster consists of the pelB and pelC genes. PelB and PelC share extensive genetic homology, with 84% amino-acid identity. The pelA/pelE genes have diverged more than the pelB/pelC genes, exhibiting 61% amino-acid identity (Tamaki et al., 1988; Barras et al., 1987). The ®rst structure of a pectate lyase, PelC, was determined from E. chrysanthemi EC16 (Yoder, Keen et al., 1993). The polypeptide chains folds in a large right-handed helical coil, forming three parallel -sheets. This novel fold is referred to as a parallel -helix. Other Pel structures reported to date include E. chrysanthemi PelE (Lietzke et al., 1994), E. chrys- anthemi PelC bound to a pentagalacturonate substrate (Scavetta et al., 1999) and the pectate lyase from Bacillus # 2002 International Union of Crystallography subtilis (Pickersgill et al., 1994). The structure of PelC bound Printed in Denmark ± all rights reserved to a pectate fragment has been particularly enlightening in 1008 Thomas et al. Pectate lyase A Acta Cryst. (2002). D58, 1008±1015 research papers Table 1 Table 2 Crystal parameters for PelA C2 and R3 crystal forms. Statistics for X-ray data collection for the C2 and R3 crystal forms. Space group C2 R3 Values in parentheses indicate the highest resolution shell. Highest resolution range for the C2 data is 1.80±1.88 AÊ and for the R3 data is 2.10±2.20 AÊ . Unit-cell parameters (AÊ , ) a = 87.549 a = 98.055 b = 53.802 b = 98.055 C2 R3 c = 71.039 c = 217.162 Ê = 109.448 = 120 Wavelength (A) 0.9795 1.543 Unit-cell volume (AÊ 3) 3.1 105 1.8 106 Temperature (K) 110 153 Ê 3 1 Â Â Resolution (AÊ ) 1.8 2.1 VM (A Da ) 2.03 2.54 Molecules per asymmetric unit 1 2 Re¯ections collected 88736 111710 Estimated solvent content (%) 29.9 44.0 Unique re¯ections 36568 44138 Diffraction limit (AÊ ) 1.6 2.1 Completeness (%) 88.3 (64.9) 98.3 (99.8) I/(I) 21.2 (6.28) 12.6 (3.5) Ih >3(iI)² (%) 90.5 (78.3) 81.1 (50.3) Rsym³ (%) 4.1 (16.0) 7.7 (31.0) postulating a catalytic role for an Arg in the enzymatic Mosaicity () 0.35 0.64 mechanism (Herron et al., 2000). ² Percentage of data with I >3(I). ³ Rsym = Ihkl Ihkl = Ihkl (single The oligosaccharide end products of PelA catalysis on a measurements excluded). h i polygalacturonate substrate vary in length from dimers to P P octamers and nonamers. In contrast, the end products of PelC and PelB are predominantly trimeric saccharides, while the 2.2. Data collection and refinement PelE end products are primarily dimeric (Preston et al., 1992). It is not clear from comparing the known structures of PelC Data-collection statistics are given in Table 2. The R3 crystal and PelE why this end-product difference occurs. data were collected on a MAR300 phosphoimaging-plate From biochemical studies, it is clear that PelA has several detector mounted on a rotating-anode generator operating at unique features relative to the other isozymes that warrant a 50 kV and 100 mA with a Supper graphite monochromator. structural characterization. PelA is the only acidic isozyme, The loop-mounted crystal was ¯ash-cooled after a serial with a pI of 4.6. While the speci®c activity of PelA is lower but transfer to a cryosolution of 2 M Li2SO4, 40% PEG 8K similar to that of PelE (Barras et al., 1987), PelA and the containing increasing percentages of ethylene glycol to a ®nal other Pel isozymes exhibit signi®cant differences in tissue- concentration of 25%. X-ray data were collected at 153 K with maceration ability. PelB, PelC and PelE cleave poly- an exposure time of 6 min, an oscillation range of 1 and a galacturonate to small oligosaccharide end products. In crystal-to-detector distance of 150 mm. contrast, PelA end products are considerably longer and more The C2 crystal data were collected with synchrotron varied (Preston et al., 1992). The structure determination of radiation at NSLS beamline X8C. Single-wavelength X-ray PelA was initiated to elucidate possible structural rationali- data were collected at 0.9795 AÊ on an ADSC Quantum4 zations in order to address these biochemical observations. detector with a crystal-to-detector distance of 150 mm. The oscillation range was 1 per frame, with an exposure time of 60 s. Data were collected at 110 K from a ¯ash-cooled crystal after a serial transfer to a cryosolution of 10% PEG 8K, 4% 2. Materials and methods PEG 1K containing increasing percentages of ethylene glycol to a ®nal concentration of 20%. 2.1. Protein purification and crystallization X-ray diffraction images were indexed and scaled with HKL The protein was puri®ed and crystallized as described (Otwinowski & Minor, 1997). The initial structure was deter- previously (Doan et al., 2000). Escherichia coli cells, which mined by molecular replacement as implemented in AMoRe contain pPEL812 containing a high-expression plasmid (Navaza, 1994). The molecular-replacement search model was construct with the pelA gene from E. chrysanthemi EC16 based on PelE residues 30±61, 70±100, 105±251, 258±280, (Tamaki et al., 1988), were grown in LB media and the protein 293±314 and 323±332. Positional parameters were optimized was puri®ed from the periplasmic fraction. The protein was by rigid-body re®nement against the C2 data. Missing portions suf®ciently pure for crystallization after DEAE ion-exchange of the PelA model were manually built into electron-density chromatography followed by hydrophobic interaction chro- maps using the computer-graphics program O (Jones et al., matography on phenyl Sepharose. During the course of the 1991). The models were checked using simulated-annealing study, two additional crystal forms of PelA were found. The C2 OMIT maps and careful monitoring of Rfree statistics form crystallizes under the same conditions as a previously (BruÈnger, 1993) as calculated by CNS (BruÈnger et al., 1998). described P21 form (Doan et al., 2000). The R3 crystal form The re®ned model from the C2 data consists of all residues in was grown at room temperature using sitting-drop vapor- the protein (1±361), 244 water molecules and one sulfate ion. diffusion techniques. A 0.006 ml solution containing 6.0 A280 The side chains of residues Glu2, Lys135, Asp137 and Lys334 of PelA in 10 mM HEPES pH 7.0 with 0.5 M Li2SO4 and 1% were modeled as Ala owing to a lack of detectable electron polyethylene glycol (PEG) 8K was equilibrated over a 0.8 ml density beyond the C atom. One residue, Tyr247, has an reservoir of 1 M Li2SO4 and 2% PEG 8K. Crystal parameters unfavorable combination of , ' backbone torsion angles as for the two new crystal forms are given in Table 1. determined by PROCHECK (Laskowski et al., 1992) (Fig. 1a). Acta Cryst. (2002). D58, 1008±1015 Thomas et al. Pectate lyase A 1009 research papers The R3 X-ray data were twinned; the twin law was deter- Table 3 mined by the Merohedral Crystal Twinning Server for partial Final re®nement statistics as a function of resolution.