Expression, purification, crystallization and preliminary X-ray analysis of the polysaccharide lyase RB5312 from the marine planctomycete Rhodopirellula baltica

This study describes the crystallization and preliminary X-ray analysis of the family PL1 polysaccharide lyase RB5312 from the marine bacterium R. baltica. Purified recombinant protein was crystallized; the crystals belonged to space group P2₁2₁2₁ and diffracted X-rays to a resolution of 1.8 Å.

Abstract

Polysaccharide lyases belonging to family PL1 act on pectins. These anionic polymers are usually produced by terrestrial plants and therefore pectinolytic enzymes are not frequently observed in marine microorganisms. The protein RB5312 from the marine bacterium Rhodopirellula baltica is distantly related to family PL1 pectate lyases, but its exact function is unclear. In this study, the expression and purification of a recombinant form of RB5312 are described. This protein was crystallized using the hanging-drop vapour-diffusion method. The crystals belongs to space group P2₁2₁2₁, with unit-cell parameters a = 39.05, b = 144.05, c = 153.97 Å, α = β = γ = 90°. A complete data set was collected to 1.8 Å resolution from a native crystal.

1. Introduction

The cell walls of marine algae and seagrasses are characterized by an abundance of anionic polysaccharides formed from carboxylic or sulfated sugars (Aquino et al., 2005 ▶; Kloareg & Quatrano, 1988 ▶). These polymers constitute a crucial carbon source for a number of marine bacteria that secrete specific glycoside hydrolases or polysaccharide lyases (Michel et al., 2006 ▶). Among these microorganisms, the Planctomycetes are recognized as one of the key phyla catalyzing important transformations in the global carbon cycle in the sea (Alldredge, 2000 ▶; DeLong et al., 1993 ▶). Planctomycetes frequently inhabit phytodetrial macroaggregates in marine environments where they mineralize this organic matter, which is mainly composed of polysaccharides (Miskin et al., 1999 ▶; Neef et al., 1998 ▶; Wang et al., 2002 ▶). Rhodopirellula baltica is a marine representative of this important bacterial phylum and it was the first to have its genome completely sequenced (Glöckner et al., 2003 ▶). Interestingly, its genome revealed an unprecedented large number of sulfatases, accompanied by numerous polysaccharidases. These observations suggest that R. baltica is able to degrade a wide range of anionic polysaccharides from marine algae and seagrasses.

Among the various enzymes found in the R. baltica genome, two have been annotated as putative pectate lyases: RB5312 (GenBank accession No. CAD74167) and RB5316 (CAD74169) (Glöckner et al., 2003 ▶). Both proteins have been assigned to family PL1 of the polysaccharide lyases by the Carbohydrate Active Enzyme website (http://www.cazy.org/; Coutinho & Henrissat, 1999 ▶). Pectins are a group of plant polysaccharides that are mainly constituted of d-­galacturonic acid (GalA) and three polysaccharide domains are often found: homogalacturonan (HGA), rhamnogalacturonan-I (RGI) and rhamnogalacturonan-II (RGII) (Mohnen, 1999 ▶; O’Neill et al., 1990 ▶). HGA is a linear homopolymer of (1,4)-α-linked d-galacturonic acid and is deposited in the cell wall of land plants in a form that has 70–80% of GalA residues methyl-esterified at the C-6 carboxyl (Mohnen, 1999 ▶; O’Neill et al., 1990 ▶). Pectate and pectin lyases cleave the α-(1,4)-linkages in HGA domains by a β-elimination mechanism, releasing an unsaturated C4—C5 bond at the non­reducing end of the cleaved polysaccharide. Pectate lyases (EC 4.2.2.2) are specific for demethylated forms of HGA and require Ca²⁺ for activity (Pilnik & Rombouts, 1981 ▶; Scavetta et al., 1999 ▶). In contrast, pectin lyases (EC 4.2.2.10) cleave methylated HGA according to a calcium-independent mechanism (Mayans et al., 1997 ▶; Pilnik & Rombouts, 1981 ▶).

The presence of putative pectinolytic enzymes in a marine bacterium is surprising since pectins are typical of terrestrial higher plants. Nevertheless, pectin-like polymers have been identified in some marine green algae (Kloareg & Quatrano, 1988 ▶) and seagrasses (Ovodov et al., 1975 ▶). To evaluate the quality of the functional predictions of Glöckner et al. (2003 ▶), we have performed our own sequence analyses using BLAST searches against the SwissProt database. RB5316 displays 37% and 29% sequence identity to the pectate lyases PelA from Emericella nidulans and PelE from Erwinia chrysanthemi, respectively (Ho et al., 1995 ▶; Lietzke et al., 1994 ▶). Therefore, RB5316 is indeed homologous to characterized members of the PL1 family and is likely to be a pectate lyase. In contrast, for the second protein RB5312 only the N-terminal region shares some similarities to characterized proteins: the pectate lyase PelA from Em. nidulans (29% sequence identity for Ala22–Ser222, RB5312 numbering) and the noncatalytic pollen allergen AgE from Ambrosia artemisiifolia (27% sequence identity for Ala19–Asp160; Rafnar et al., 1991 ▶). The C-terminal region of RB5312 (Gly225–Pro455) displays no significant similarity to any characterized protein. At this low level of sequence identity, the exact function of RB5312 is unclear and the prediction that it is a pectate lyase is difficult to substantiate.

Therefore, the focus of our study is to unravel the biochemical function of the marine polysaccharide lyase RB5312. Determination of its crystal structure will also help to establish the role of this divergent member of the PL1 family. We have thus cloned the rb5312 gene from R. baltica into Escherichia coli and produced pure recombinant protein. We also report the crystallization of RB5312 and the preliminary X-ray analysis of the crystals.

2. Experimental

2.1. Overexpression and purification of RB5312

According to the program SignalP (Bendtsen et al., 2004 ▶), RB5312 features an N-terminal signal peptide that is cleaved between residues Ala23 and Gln24. The nucleotide sequence corresponding to the mature RB5312 protein (Gln24–Pro455) was amplified by PCR from R. baltica genomic DNA using a set of primers (see Table 1 ▶) at a half-hybridization temperature of 345 K. The obtained PCR product was then purified using the Qiagen QIAquick system and digested with BglII/EcoRI (5′/3′ ends) in NEB2 buffer (BioLabs) at 310 K for 3 h. Ligation was performed overnight at 293 K using T4 ligase (Sigma) in a pFO4 vector (derivative of pET15) pre-digested with BamHI/EcoRI. This resulted in a gene coding for a recombinant protein (441 residues in total, 48 701 Da) flanked by an N-terminal tag encompassing a methionine followed by six histidine residues and two residues, arginine and serine, corresponding to the BglII restriction site. The ligation mixture was transformed into E. coli DH5α strain. The recombinant plasmid was extracted with the Wizard Plus SV Minipreps kit (Promega) and used to transform E. coli BL21 (DE3). The recombinant cells were incubated at 293 K in ZYP-5052 medium (Studier, 2005 ▶) with 100 µg ml⁻¹ ampicillin until saturation of the culture (final OD_600nm ≃ 15). The cells were harvested by centrifugation (4000g, 277 K, 20 min). The cell pellet was resuspended in buffer A (50 mM NaH₂PO₄ pH 8.0, 500 mM NaCl, 50 mM imidazole, 5% glycerol). The cells were disrupted using a French press and the lysate was cleared by centrifugation (50 000g, 277 K, 30 min). The protein supernatant was applied onto a 10 ml column consisting of IMAC HyperCell resin (Pall Corporation) charged with 100 mM NiSO₄ and pre-equilibrated with buffer A. After a step of washing with buffer A (two column volumes), the protein was eluted with a 60 ml linear gradient from buffer A to buffer B (50 mM NaH₂PO₄ pH 8.0, 500 mM NaCl, 400 mM imidazole, 5% glycerol) at a flow rate of 1 ml min⁻¹. The protein eluted at 160 mM imidazole. The protein peak was analyzed by SDS–PAGE (Fig. 1 ▶) and the purest fractions were pooled (10 ml final volume). The volume was reduced to 4 ml by ultrafiltration on an Amicon membrane (polyethersulfone, 30 kDa cutoff). The protein sample was injected onto a Superdex 75 HiLoad (GE Healthcare) column pre-equilibated with buffer C (50 mM HEPES–HCl pH 7.5, 50 mM NaCl). Elution was performed with 60 ml buffer C at a flow rate of 1 ml min⁻¹. The final protein yield was 2 mg per 500 ml of culture medium. The purified protein was concentrated to 9.7 mg ml⁻¹ by ultrafiltration on an Amicon membrane.

Table 1. Oligonucleotides used for the cloning of rb5312 .

The primers were designed for a half-hybridization temperature of 345 K.

Forward	GGGGGGAGATCTCAGAAGCCATTGGCCTTTCCGAC
Reverse	CCCCCCGAATTCTTACGGGATGCTGTTGATGTATTGTTC

Figure 1.

SDS–PAGE gel (12.5%) stained with Coomassie Blue showing the different steps in the heterologous expression and purification of the polysaccharide lyase RB5132. Lane 1, molecular-weight markers (kDa); lane 2, soluble fraction of nontransformed E. coli BL21 lysate; lane 3, soluble fraction of E. coli lysate with expression; lane 4, fraction with maximum intensity (absorption at 280 nm) after Ni–IMAC column chromatography; lane 5, fraction with maximum intensity after size-exclusion column chromatography.

2.2. Preliminary pectinolytic activity tests

The pectinolytic activity of RB5312 was tested spectrophotometrically at 235 nm with a UV-2041 PC spectrophotometer (Shimadzu, Japan) by measuring the increase in absorbance caused by the formation of C4–C5 unsaturated sugars. Measurements were taken every 25 s over 10 min at 313 K. The standard assay (1 ml) contained 5 µl purified enzyme at 1 mg ml⁻¹, 995 µl 100 mM Tris–HCl pH 8.0 and 1%(w/v) polygalacturonic acid (PGA) from citrus fruit (Sigma). This activity test was also performed with 995 µl 100 mM Tris–HCl pH 8.0, 1%(w/v) PGA and 5 mM EGTA as a control.

2.3. Protein crystallization

All crystallization experiments were carried out at 292 K. Initial crystallization trials were performed with the Wizard I and II (Emerald BioStructures, Inc.) and JCSG+ Suite (Qiagen) kits; that is, a total of 192 conditions in two 96-well plates from Corning. The trials were set up using a Cartesian crystallization robot and the sitting drops were made up by mixing 300 nl protein solution (9.7 mg ml⁻¹ in 50 mM HEPES buffer pH 7.5, 50 mM NaCl) with 150 nl reservoir solution. Subsequently, the best conditions were optimized in 24-well Linbro plates using the hanging-drop vapour-diffusion method. These drops were prepared on siliconized cover slips by mixing 2 µl protein solution with 1 µl well solution. The drops were equilibrated against 0.5 ml reservoir solution.

2.4. Data collection and X-ray diffraction analysis

X-ray diffraction data were collected from a crystal of recombinant RB5312 at 100 K on beamline ID14-EH2 at the ESRF (Grenoble, France) using an ADSC Quantum 4R CCD detector. All crystals were flash-cooled in a liquid-nitrogen stream. Since the crystallization condition contained 40% 2-methyl-2,4-pentanediol (MPD), it was not necessary to add a cryoprotectant. The wavelength of the synchrotron X-rays was 0.933 Å. The crystal was rotated through 100° with 0.4° oscillation per frame. Further data-collection statistics are given in Table 2 ▶. All raw data were processed using the program MOSFLM (Leslie, 1992 ▶). The resultant data were merged and scaled using the program SCALA, which is part of the CCP4 suite (Collaborative Computational Project, Number 4, 1994 ▶). An attempt to solve the structure of RB5312 by molecular replacement was performed using the program AMoRe (Navaza, 2001 ▶). The model was selected by a sequence search using BLAST against the PDB sequence database.

Table 2. Data-collection statistics of RB5312 crystals.

Values in parentheses are for the highest resolution shell.

Space group	P212121
Unit-cell parameters (Å, °)	a = 39.05, b = 144.05, c = 153.97, α = β = γ = 90
Resolution range (Å)	52.63–1.8 (1.90–1.80)
No. of observations	304224 (14846)
No. of unique reflections	76 671 (7858)
Completeness (%)	93.3 (67.0)
〈I/σ(I)〉	17.3 (3.9)
Redundancy	4.0 (1.9)
Rmerge† (%)	5.4 (16.7)

^† R _merge = .

3. Results

RB5312 was expressed in E. coli without its signal peptide as a soluble protein. This recombinant protein was purified by a combination of IMAC and size-exclusion chromatography in sufficient quantities for crystallization (Fig. 1 ▶). A linear increase in A _235nm was observed when RB5312 was added to a reaction assay containing polygalacturonic acid from citrus fruit. This phenomenon is completely suppressed by the addition of 5 mM EGTA. Therefore, RB5312 displays pectinolytic activity and its mechanism is calcium-dependent, suggesting a pectate-lyase specificity. To confirm this substrate specificity, purification of the terminal products of RB5312 is under way. The purified oligosaccharide will be characterized by ¹H NMR (Dabin et al., in preparation).

The crystallization screening of RB5312 resulted in the identification of several successful conditions containing MPD and polyethylene glycol (PEG). The optimized conditions for crystal growth are 40% MPD, 4% PEG 8000, 150 mM sodium cacodylate pH 6.0 at 292 K. Needle-shaped crystals grew within several days (Fig. 2 ▶). Most crystals of the recombinant native protein diffracted to 2.5 Å resolution. However, by screening numerous crystals (∼15) one single crystal (of dimensions 0.3 × 0.05 × 0.05 mm) was found to diffract to 1.8 Å resolution. A complete data set was collected from this crystal and the data-collection quality is reported in Table 2 ▶. The space group was determined to be P2₁2₁2₁, with unit-cell parameters a = 39.05, b = 144.05, c = 153.97 Å, α = β = γ = 90°. The asymmetric unit most probably contains two molecules, giving a crystal volume per protein weight (V _M) of 2.17 ų Da⁻¹ and a solvent content of 43% by volume (Matthews, 1968 ▶).

Figure 2.

Needle-shaped crystals of polysaccharide lyase RB5132.

The closest sequence match to RB5312 in the PDB is Jun a 1 (PDB code 1pxz), the major cedar pollen allergen from Juniperus ashei (Czerwinski et al., 2005 ▶). Jun a 1 adopts a β-helix fold and clearly belongs to the PL1 family, but its sequence identity to RB5312 is hardly significant (only 14%). Although the number of insertions and deletions between the two sequences is low, we nevertheless attempted to solve the structure of RB5312 by molecular replacement using the atomic coordinates of 1pxz. However, as expected, these attempts failed. We have therefore decided to produce seleno­methionine-labelled protein and the first crystallization trials are under way.