Crystallization and preliminary structural characterization of the two actin-depolymerization factors of the malaria parasite

The expression, purification and crystallization of Plasmodium actin-depolymerization factors 1 and 2 are described. X-ray diffraction data were collected to 2.0 and 2.1 Å resolution, respectively, and the structures of both proteins in solution were characterized.

Abstract

The malaria parasite Plasmodium depends on its actin-based motor system for motility and host-cell invasion. Actin-depolymerization factors are important regulatory proteins that affect the rate of actin turnover. Plasmodium has two actin-depolymerization factors which seem to have different functions and display low sequence homology to the higher eukaryotic family members. Plasmodium actin-depolymerization factors 1 and 2 have been crystallized. The crystals diffracted X-rays to maximum resolutions of 2.0 and 2.1 Å and belonged to space groups P3₁21 or P3₂21, with unit-cell parameters a = b = 68.8, c = 76.0 Å, and P2₁2₁2, with unit-cell parameters a = 111.6, b = 57.9, c = 40.5 Å, respectively, indicating the presence of one or two molecules per asymmetric unit in both cases.

1. Introduction

Malaria is one of the most devastating health threats to mankind and particularly affects developing countries in (sub)tropical regions. Malaria is caused by apicomplexan parasites of the genus Plasmodium. These parasites utilize a unique actin-based motor system for gliding motility and host-cell invasion (Sibley, 2004 ▶). Rapid polymerization and depolymerization of actin, which is regulated by a rather small set of actin filament-binding and monomer-binding proteins, is a prerequisite for movement of the parasite (Schüler & Matuschewski, 2006 ▶). Plasmodium actin filaments are extremely short and are located between the plasma membrane and the inner membrane complex of the parasite (Schatten et al., 2003 ▶; Schmitz et al., 2005 ▶). They are linked to host-cell surface receptors by a transmembrane protein complex, TRAP-aldolase (Jewett & Sibley, 2003 ▶), and to an array of unconventional myosins immobilized in the inner membrane complex (Bergman et al., 2003 ▶).

Actin-depolymerization factors (ADFs) belong to the ADF/cofilin family, which comprises a group of important regulators of actin-filament dynamics (van Troys et al., 2008 ▶; Bernstein & Bamburg, 2010 ▶). On the one hand ADFs bind to the side of actin filaments, thereby destabilizing the monomer–monomer interactions within the filament and leading to accelerated depolymerization rates. On the other they bind to ADP-bound actin monomers, inhibiting nucleotide exchange to ATP and thereby decreasing the rate of polymerization (Carlier et al., 1997 ▶; Bamburg, 1999 ▶; Paavilainen et al., 2008 ▶). At least two of the apicomplexan ADFs have been shown to be unable to bind to actin filaments (Schüler et al., 2005 ▶; Mehta & Sibley, 2010 ▶). Unlike other Apicomplexa, Plasmodium species have two ADFs. Of these, ADF1 is similar to the ADFs present in other parasites of this class, while ADF2 more closely resembles conventional ADFs at the sequence level (Schüler et al., 2005 ▶).

Because of constantly increasing drug resistance and the lack of an effective vaccination against malaria, there is an enormous need for new targets for drug or vaccine development. The actin cytoskeleton of the malaria parasite is an attractive target for pharmaceutical design because it is essential for the survival of the parasite and the invasion of host cells and is also significantly different from the cytoskeleton of higher eukaryotes at the protein level. Here, we describe the expression, purification, crystallization and preliminary structural characterization of P. falciparum ADF1 (PfADF1; gi:23613057; PlasmoDB PFE0165w) and P. berghei ADF2 (PbADF2; gi:68159351; PlasmoDB PB001373).

2. Materials and methods

2.1. Cloning, expression and purification of the proteins

The coding regions of PfADF1 and PbADF2 were amplified by PCR on total genomic DNA from P. falciparum and P. berghei using the following primers with BamHI and XhoI sites at the 5′ and 3′ ends, respectively: PfADF1 forward, 5′-CGTGGATCCGGGG­TCAAGG; reverse, 5′-CCGCTCGAGTGTGTTTAATATTATAGT; PbADF2 forward, 5′-AGTTGGATCCATGGTTTCAGGAG; reverse, 5′-AG­GACTCGAGTTATGTGTTTAAAATAATAG. The PCR products were digested with the respective enzymes and ligated into the pGEX-6P-1 vector (GE Healthcare). The resulting constructs were verified by restriction digestion followed by sequencing. Recombinant glutathione S-transferase (GST)-PfADF1 was expressed in Escherichia coli BL21 (DE3) RIPL cells. Transformed cells were grown for 36 h at 291 K in 800 ml autoinduction medium (Studier, 2005 ▶) supplemented with 100 µg ml⁻¹ ampicillin and 25 µg ml⁻¹ chloramphenicol. After 24 h, an additional dose of 125 µg ml⁻¹ ampicillin, 0.03% glucose, 0.13% lactose and 0.3% glycerol was added to the culture to maintain cell growth and plasmid stability. The cells were harvested by centrifugation at 277 K and 4000g for 40 min and frozen; they were sub­sequently resuspended in lysis buffer consisting of 20 mM Tris–HCl pH 8.5, 50 mM NaCl, 5% glycerol, 10 mM dithiothreitol (DTT) and disrupted by sonication. The fusion protein was bound to glutathione Sepharose (GE Healthcare) in gravity-flow columns and eluted using a buffer consisting of 20 mM Tris–HCl pH 8.5, 50 mM NaCl, 5% glycerol, 10 mM DTT, 10 mM reduced l-glutathione. The pooled fractions containing the fusion protein were dialyzed overnight against 20 mM 4-(2-hydroxyethyl)-1-­piperazineethanesulfonic acid (HEPES) pH 7.1, 2.5 mM DTT, after which the N-terminal GST tag was cleaved using 4.5 µg ml⁻¹ recombinant 3C protease for 24–36 h at 277 K. The cleaved PfADF1 contained five additional residues (GPLGS) before the N-terminal Met. GST and other remaining contaminants were bound to Q-­Sepharose FF (GE Healthcare) and the unbound PfADF1 was concentrated; final purification was performed by size-exclusion chromatography using a HiLoad 16/60 Superdex 75 pg column (GE Healthcare) equilibrated with 20 mM 3-(N-morpho­lino)propanesulfonic acid (MOPS) pH 6.5, 50 mM NaCl, 5%(v/v) glycerol, 2 mM DTT.

Recombinant GST-PbADF2 protein was expressed in E. coli BL21 Codon Plus (DE3) Rosetta cells grown in Luria–Bertani medium (1% tryptone, 0.5% yeast extract, 1% NaCl). Expression was induced with 0.6 mM isopropyl β-d-1-thiogalactopyranoside for 12 h at 291 K. The bacteria were harvested by centrifugation at 277 K and 15 300g for 15 min and frozen. The frozen cells were resuspended in lysis buffer containing 50 mM NaCl, 10% glycerol, 10 mM Tris–HCl pH 8.0 supplemented with an EDTA-free protease-inhibitor cocktail (Roche Diagnostics GmbH) and disrupted by sonication. The fusion protein was bound to glutathione Sepharose (GE Healthcare) in a gravity-flow column, eluted using a buffer containing 300 mM NaCl, 10% glycerol, 5 mM DTT, 20 mM reduced l-glutathione, 10 mM Tris–HCl pH 8.0 and dialyzed overnight against the same buffer without glutathione. The GST tag was cleaved using 3.5 µg ml⁻¹ recombinant 3C protease for 6–8 h at 277 K. PbADF2 also contained the sequence GPLGS before the N-terminal Met after cleavage. The final purification step was performed by size-exclusion chromatography in buffer consisting of 300 mM NaCl, 10% glycerol, 5 mM DTT and 10 mM Tris–HCl pH 8.0 using a HiLoad 16/60 Superdex 75 pg column (GE Healthcare). Pure PbADF2 was flash-frozen in liquid nitrogen and stored at 193 K.

2.2. Characterization of purified PfADF1 and PbADF2

The identity of the purified proteins was verified by tryptic peptide mapping using mass spectrometry at the Biocenter Oulu Proteomics Core Facility. Because PfADF1 had a tendency to form both soluble and insoluble aggregates after cleavage of the GST tag, a thermal shift assay (Ericsson et al., 2006 ▶) using the FRET channel in a Bio-Rad CFX96 Real Time System and a C1000 Thermal Cycler was performed to determine the optimal buffer conditions for purification and crystallization. Several buffers with different pH values and salts were tested. Each well contained 40 µl buffer, 8 µg protein and 4 µl 12× Sypro Orange dye (Invitrogen) in a total volume of 48 µl. The samples were heated from 291 to 368 K at a rate of 1 K min⁻¹ in 96-­well PCR plates sealed with optical sealing tape.

Analytical gel filtration injecting 100 µg PfADF1 in 20 mM MOPS pH 6.5, 100 mM NaCl, 2 mM DTT at a flow rate of 0.2 ml min⁻¹ onto a Superdex 75pg 10/300 GL column (GE Healthcare) combined inline with three-angle static and dynamic light-scattering detectors (Wyatt Technology) was performed to determine the oligomeric state and monodispersity of PfADF1.

Synchrotron-radiation circular-dichroism spectroscopy (SRCD) was used to assess the folding and secondary-structure content of PbADF2. For the measurements, two samples, one of freshly purified protein and the second of protein stored unfrozen on ice for two weeks, at a concentration of 1 mg ml⁻¹ were dialyzed against 10 mM potassium phosphate buffer pH 7.7, 50 mM NaF, 10% glycerol. The SRCD spectra were measured on beamline 3m-NIM at BESSY, Berlin, Germany in a cuvette with a 0.01 cm path length at 298 K from 170 to 280 nm wavelength with a 1 nm interval, as described previously (Majava et al., 2009 ▶).

Small-angle X-ray scattering (SAXS) data for PfADF1 were measured on the MAX-Lab I711 beamline using PfADF1 at a con­centration of 4.0 mg ml⁻¹ in 20 mM MOPS pH 6.5, 100 mM NaCl, 5%(v/v) glycerol, 2 mM DTT. SAXS data for PbADF2 were measured on EMBL/DESY beamline X33 using PbADF2 at a concentration of 5 mg ml⁻¹ in 300 mM NaCl, 10% glycerol, 5 mM DTT and 10 mM Tris–HCl pH 8.0. The corresponding buffer was always used for a blank experiment. The data were analyzed using programs from the ATSAS package (Konarev et al., 2006 ▶). Further processing of the data was performed using PRIMUS (Konarev et al., 2003 ▶) by subtracting the scattering by the buffer and normalizing the scattering by the proteins to a concentration of 1 mg ml⁻¹. The molecular weights of PfADF1 and PbADF2 were estimated by comparing the forward scattering I(0) with that of a standard solution of human calmodulin or mouse 2′,3′-cyclic nucleotide 3′-phosphodiesterase. The distance distributions were obtained using GNOM (Svergun, 1992 ▶) and further used for ab initio modelling in DAMMIN (Svergun, 1999 ▶). An averaged model was generated from several runs using programs from the DAMAVER package (Volkov & Svergun, 2003 ▶). An estimate of the molecular weight was also calculated based on the model volume compared with those of standard proteins.

2.3. Crystallization

For crystallization, purified PfADF1 and PbADF2 were concentrated in the final purification buffers (20 mM MOPS pH 6.5, 50 mM NaCl, 5%(v/v) glycerol, 2 mM DTT for PfADF1 and 10 mM Tris–HCl pH 8.0, 300 mM NaCl, 10% glycerol, 5 mM DTT for PbADF2) to approximately 10 mg ml⁻¹ using centrifugal concentrators with a molecular-weight cutoff of 5 kDa (Millipore). Crystallization screens were set up in MRC conical-bottom sitting-drop vapour-diffusion plates (Molecular Dimensions) both at room temperature and at 277 K, mixing 0.5 µl of protein and well solution. The reservoir solution volume was 50 µl. PfADF1 crystals (Fig. 1 ▶ a) formed in 10 d at 277 K in 2.5 M ammonium sulfate, 0.1 M HEPES pH 7.6–8.0. The crystal used for data collection was grown at pH 7.8. PbADF2 crystals (Fig. 1 ▶ b) grew overnight at both temperatures in 24–28% polyethylene glycol (PEG) 10 000, 0.1 M sodium acetate pH 4.5–5.5. A crystal grown at 277 K in 26% PEG 10 000, 0.1 M sodium acetate pH 5.2 was used for data collection.

Figure 1.

(a) A single crystal of PfADF1 grown in 2.5 M ammonium sulfate, 0.1 M HEPES pH 7.8. (b) A single crystal of PbADF2 grown in 26% PEG 10 000, 0.1 M sodium acetate pH 5.2. The maximum dimension of both crystals was approximately 200 µm.

2.4. Diffraction data collection

Prior to data collection, crystals were flash-frozen in liquid nitrogen without additional cryoprotectants. Because the PfADF1 crystal suffered on freezing, as was apparent from the smeary diffraction spots and several strong ice rings, annealing by blocking the stream of liquid nitrogen for 10 s was performed to improve the diffraction properties of the crystal. Subsequently, diffraction data sets were collected from single crystals of PfADF1 and PbADF2 to 2.0 and 2.1 Å resolution, respectively, using MAR 165 and MAR Mosaic225 CCD detectors mounted on beamlines I911-2 and I911-3 at MAX-Lab, Lund, Sweden. The data were indexed and integrated using XDS (Kabsch, 1993 ▶) and scaled using XSCALE (Kabsch, 1993 ▶) using the graphical XDSi interface (Kursula, 2004 ▶). The complete data-collection statistics are shown in Table 1 ▶.

Table 1. Data-collection statistics for PfADF1 and PbADF2.

Values in parentheses are for the highest resolution shell.

	PfADF1	PbADF2
Beamline	I911-2	I911-3
Wavelength (Å)	1.0380	0.9790
Beam size (mm)	0.3 × 0.3	0.15 × 0.25
Space group	P3121/P3221	P21212
Crystal dimensions (mm)	0.2 × 0.15 × 0.15	0.2 × 0.05 × 0.05
Unit-cell parameters	a = b = 68.84, c = 76.03, α = β = 90, γ = 120	a = 111.58, b = 57.90, c = 40.49, α = β = γ = 90
Resolution range (Å)	20.0–2.0 (2.05–2.0)	20–2.1 (2.15–2.1)
Oscillation per frame (°)	1	0.25
No. of frames collected	180	653
No. of measured reflections	145966 (7050)	97420 (7461)
No. of unique reflections	14010 (1004)	15859 (1168)
Criterion for observed reflections [I/σ(I)]	−3	−3
Multiplicity	10.4 (7.0)	6.1 (6.4)
Completeness (%)	96.5 (94.5)	99.1 (99.8)
Average I/σ(I)	38.5 (14.5)	16.4 (4.2)
Rmerge† (%)	4.6 (27.5)	6.0 (64.9)

^† R _merge = , where I _i(hkl) is the intensity of the ith measurement of reflection hkl and 〈I(hkl)〉 is the mean intensity of reflection hkl.

3. Results and discussion

PfADF1 and PbADF2 were purified using affinity and size-exclusion chromatography. For PfADF1, an additional ion-exchange chromatography step had to be added between these steps in order to avoid precipitation during concentration before size-exclusion chroma­tography. A thermal shift assay was used to determine the stability of PfADF1 in various buffers. The protein was most stable at pH 8.5 or 6.5 with low salt concentrations, such as 50 mM NaCl and 0–5% glycerol. The apparent molecular weights of PfADF1 and PbADF2 as deduced from SDS–PAGE (Laemmli, 1970 ▶) were, as expected, 14 and 17 kDa, respectively, and analytical gel filtration together with static and dynamic light scattering confirmed PfADF1 to be monomeric in solution (data not shown). The molecular weight of PfADF1 deduced from static light scattering was 13.5 kDa, which was very close to the calculated molecular weight of 13.7 kDa, and dynamic light scattering showed the protein to be monodisperse. The SRCD data indicated that PbADF2 was folded and had the expected secondary-structure content of a member of the ADF/cofilin family (data not shown). It could also be seen that PbADF2 partially loses its secondary structure when stored unfrozen on ice for a period of two weeks.

We determined the solution structures of PfADF1 and PbADF2 using SAXS. These structures also show that both proteins are folded and monomeric in solution. It can also be seen that PbADF2 has a more elongated structure, as expected from the structures of other members of the ADF/cofilin family, whereas PfADF1 is more spheri­cal (Fig. 2 ▶).

Figure 2.

Calculated SAXS envelopes of (a) PfADF1 and (b) PbADF2 and the corresponding X-ray scattering curves. The dots are the measured scattering intensities and the solid line describes the fit of the calculated model to the measured data.

Crystallization trials were performed using different protein concentrations and numerous commercially available and homemade crystallization screens both manually and using robotics. Finally, single crystals sufficiently large for diffraction data collection grew in manual sitting-drop vapour-diffusion setups at 277 K. The PfADF1 crystal belonged to space group P3₁21 or P3₂21, with unit-cell parameters a = b = 68.8, c = 76.0 Å (Table 1 ▶). The Matthews coefficient (V _M; Matthews, 1968 ▶) of 3.71 or 1.86 ų Da⁻¹ indicated the presence of one or two monomers in the asymmetric unit, with possible solvent contents of 67 or 34%, respectively. The crystal of PbADF2 belonged to space group P2₁2₁2, with unit-cell parameters a = 111.6, b = 57.9, c = 40.5 Å (Table 1 ▶). The V _M of 4.39 or 2.19 ų Da⁻¹ indicated the presence of one or two monomers in the asymmetric unit, with possible solvent contents of 72 or 44%, respectively. The lack of significant peaks above the noise level in addition to the peaks corresponding to the crystallographic symmetry in a self-rotation function calculated using MOLREP (Vagin & Teplyakov, 1997 ▶) for both data sets indicates either the presence of only one molecule in the asymmetric unit or that a possible noncrystallographic symmetry axis is parallel with one of the crystallographic symmetry axes. As discussed above, both PfADF1 and PbADF2 were shown to be monomeric in solution using several methods.

High-resolution diffraction data sets could be collected from both PfADF1 and PbADF2 crystals. The data sets could be processed to 2.0 and 2.1 Å resolution, respectively, with good statistics (Table 1 ▶). For the PfADF1 crystal, it seems likely that higher resolution data could possibly be obtained. The PfADF1 crystal initially suffered on freezing without cryoprotection, but the strong ice rings could be removed and the diffraction limit and the quality of the diffraction spots could be clearly improved by annealing. Structure determination of both apicomplexan ADFs by molecular replacement is on­going. Several crystal and solution structures of members of the ADF/cofilin family from plants (Bowman et al., 2000 ▶), yeast (Andrianantoandro & Pollard, 2006 ▶), amoebae (Blanchoin & Pollard, 1998 ▶) and mammals (Pope et al., 2004 ▶), as well as of complexes with monomeric actin (Paavilainen et al., 2008 ▶), are available in the Protein Data Bank. All of these proteins have rather low sequence identity to the apicomplexan ADFs, with the closest homologues sharing about 30% sequence identity. Based on a promising molecular-replacement solution, it seems likely that the correct space group of the PfADF1 crystal is P3₂21. From the systematic absences, it is clear that the PbADF2 crystal belongs to space group P2₁2₁2 and this also seems to be confirmed by the molecular-replacement trials. We expect the crystal structures of Plasmodium ADFs to explain the functional differences between ADFs from apicomplexan parasites and higher eukaryotes such as their mammalian hosts.