Crystallization of a novel metal-containing cupin from Acidobacterium sp. and preliminary diffraction data analysis

Recombinant AciX9_0562 from Acidobacterium sp. MP5ACTX9 containing sequence motifs characteristic of the RmlC-type cupins superfamily and containing Pfam motif PF07883 has been successfully cloned, expressed and purified, and crystallized in a number of conditions from the Morpheus protein crystallization screen.

Abstract

Recombinant AciX9_0562 from Acidobacterium sp. MP5ACTX9 (UniProt ID E8WYN5) containing sequence motifs characteristic of the RmlC-type cupins superfamily and containing Pfam motif PF07883 has been successfully cloned, expressed and purified. AciX9_0562 crystallized in a number of conditions from the Morpheus protein crystallization screen. The best crystal diffracted to 2.7 Å resolution (space group C222₁; unit-cell parameters a = 125.29, b = 254.63, c = 82.99 Å). Structure solution was facilitated by the automated molecular-replacement pipeline BALBES. The initial solution was automatically rebuilt using the PHENIX AutoBuild wizard, with final R and R _free values of 0.23 and 0.26, respectively. The structure is currently undergoing manual refinement.

1. Introduction  

The AciX9_0562 gene from Acidobacterium sp. MP5ACTX9 (Uni­Prot ID E8WYN5) encodes a 131-amino-acid protein of uncharacterized function with a molecular weight of 14.25 kDa and a calculated isoelectric point of 5.74. A BLAST search showed that the protein is highly conserved among many bacteria; however, all sequences with high identity are annotated as ‘hypothetical proteins’. An InterPro scan (Hunter et al., 2009 ▶) for sequence motifs revealed that AciX9_0562 belongs to the RmlC-type cupins superfamily (Wilson et al., 2009 ▶) and also contains Pfam motif PF07883 (Finn et al., 2010 ▶), a conserved cupin 2 motif (Dunwell, 1998 ▶). Currently, the cupin superfamily comprises 50 families consisting of members with diverse functions ranging from enzymatic activities such as dioxygenases, decarboxylases, hydrolases, isomerases and epimerases to non-enzymatic functions such as auxin-binding proteins, nuclear transcription factors and seed storage proteins (Dunwell et al., 2000 ▶; Agarwal et al., 2009 ▶). However, no functional annotation has yet been made for AciX9_0562.

‘Cupa’ is the Latin term for a small barrel. All proteins that belong to the cupins adopt a barrel-like structure (Dunwell et al., 2001 ▶, 2004 ▶). They can occur as monocupins or bicupins and form different oligomerization states (Dunwell et al., 2004 ▶). Each cupin domain typically consists of two conserved primary-sequence motifs. Each motif corresponds to two β-strands. The motifs are separated by a less conserved region which consists of a minimum of 11 amino acids (Dunwell et al., 2001 ▶).

Here, we report the successful crystallization and structure solution of AciX9_0562 from Acidobacterium sp. MP5ACTX9.

2. Material and methods  

2.1. Plasmid construction  

The synthetic gene (gene ID 322434201; see Supplementary Material1) encoding a putative cupin 2 conserved barrel domain protein (YP_004216413) from Acidobacterium sp. MP5ACTX9 was ordered codon-optimized for expression in Escherichia coli from GeneArt (Life Technologies). The gene was recloned into pEHisTEV vector via NcoI and HindIII restriction sites to introduce an N-­terminal TEV-cleavable His tag (Liu & Naismith, 2009 ▶). The pEHisTEV construct was confirmed by sequencing (LGC Genomics, Berlin, Germany) and transformed into the expression host E. coli BL21 (DE3) Gold.

2.2. Expression and purification  

E. coli BL21 (DE3) Gold cells harbouring pEHisTEV-cupin were grown in LB (lysogeny broth) medium supplemented with kanamycin (50 µg ml⁻¹) at 310 K. The expression of recombinant protein was initiated by the addition of 0.1 mM IPTG to the cultures on reaching an OD₆₀₀ of ∼0.8 and cultivation was continued at 293 K for 20 h. Protein expression and localization in cell-extract fractions was monitored by SDS–PAGE. The cells were harvested, resuspended in cold buffer A (20 mM sodium phosphate buffer pH 7.4 containing 0.5 M NaCl and 10 mM imidazole) and disrupted by sonication (Branson Sonifier S-250; 6 min, 80% duty cycle, output 7). The cell lysate was centrifuged for 1 h at 50 000g to remove unbroken cells and insoluble material. The cell-free lysate was filtered through a 0.45 µm syringe filter and incubated with Ni Sepharose 6 Fast Flow (GE Healthcare, Uppsala, Sweden) for 20 min. The Ni Sepharose was then transferred into an empty PD-10 column. Impurities were removed by washing with 30 mM imidazole followed by elution of the bound protein with 300 mM imidazole in buffer A. Fractions were analysed by SDS–PAGE, pooled and concentrated using Vivaspin 20 Centrifugal Filter Units (10 kDa molecular-weight cutoff; Sartorius) and desalted on PD-10 desalting columns (GE Healthcare) into 20 mM Tris–HCl, 200 mM NaCl pH 7. The N-terminal His tag was cleaved using 100 µg TEV protease (expressed and purified as His-tagged protein in-house) per milligram of protein in 20 mM Tris–HCl, 200 mM NaCl pH 7, 1 mM DTT, 0.5 mM EDTA at 293 K for 20 h. His tag and TEV protease were removed by reloading the protein solution onto Ni Sepharose and collecting the cleaved protein in the flowthrough. The completeness of the cleavage was analysed by SDS–­PAGE. The cleaved samples were concentrated and loaded onto either a Superdex 75 10/30 column for qualitative analysis of the protein size and oligomerization state or a Superdex 200 16/60 column for preparative purification. The purified protein was again concentrated using Vivaspin 20 Centrifugal Filter Units (10 kDa molecular-weight cutoff; Sartorius). For both columns 20 mM Tris–HCl, 200 mM NaCl pH 7 was used as the running buffer. Prior to crystallization, purified and concentrated samples were stored in a cold room at 279 K and were transported on ice when necessary.

2.3. Crystallization  

Purified protein samples with and without His tag were subjected to screening for crystallization conditions using the Morpheus crystallization screen (Molecular Dimensions; Gorrec, 2009 ▶) and the sitting-drop method in SWISSCI 3-Well crystallization plates. Screening drops of total volume 600 nl (1:1 screening solution to protein solution at about 10 mg ml⁻¹) were dispensed using an Oryx8 Protein Crystallization Robot (Douglas Instruments Ltd) at room temperature (293 K) and the plate was stored at 290 K between inspections. Protein variants with and without His tag crystallized rapidly in several conditions. Initial crystals were observed directly after screen setup. Two conditions that produced crystals of both sample variants and that contained the visually most promising crystals were selected for further optimization: H02 [10%(w/v) PEG 8000, 20%(v/v) ethylene glycol, 0.02 M each of the amino acids dl-­Ala, dl-Lys, dl-Ser, l-Glu and glycine, 0.1 M MES–imidazole buffer pH 6.5] and E05 [10%(w/v) PEG 20 000, 20%(v/v) PEG MME 550, 0.03 M each of the ethylene glycols diethylene glycol, penta­ethylene glycol, tetraethylene glycol and triethylene glycol, 0.1 M MOPS/HEPES Na buffer pH 7.5]. In both conditions initial crystals were observed as early as the day after setup and reached maximum size within a one-week period.

In order to improve the initial crystals, an optimization grid was prepared. In this grid, the original drop components were diluted with Milli-Q water: the protein sample along the long plate axis and the respective precipitant solution along the short plate axis. The resulting serial dilution grid had the following compositions in the plate corners: A01, 30% protein and 50% precipitant; A12, 50% protein and 50% precipitant (original hit condition); H01, 30% protein and 30% precipitant; H12, 50% protein and 30% precipitant. Exact drop compositions were calculated using the XStep program of the Oryx8 Protein Crystallization Robot. Two SWISSCI 2-Well crystallization plates, one for each of the optimized conditions, were dispensed with 1000 nl total drop volumes. Each well of the plate was occupied by the protein sample variants and was equilibrated against 1 M NaCl solution (Newman, 2005 ▶). The best diffracting crystals for the sample without His tag were obtained from modified condition H02 with the following composition: 33.5% protein solution and 35.5% precipitant solution (Fig. 1 ▶). The crystals were vitrified without cryoprotection and the best data-set parameters are reported.

Figure 1.

Crystals of AciX9_0562 from Acidobacterium sp. MP5ACTX9 before (a–c) and after (d) optimization (optimized Morpheus screen condition H02, crystal radius of between 0.05 and 0.10 mm).

The sample with the His tag did not produce crystals of sufficient quality for structure determination.

2.4. Structure determination  

The best native data were collected on beamline ID23-2 at the European Synchrotron Radiation Facility (ESRF), Grenoble, France at 100 K using a MAR 225 detector (Fig. 2 ▶). 180 images collected with a 1.0° oscillation range were processed using the program MOSFLM, imported into the CCP4 program suite using the program POINTLESS (Winn et al., 2011 ▶; Evans, 2006 ▶) and scaled using the program SCALA (Evans, 2006 ▶).

Figure 2.

Diffraction image of a crystal of AciX9_0562 from Acidobacterium sp. MP5ACTX9 collected on the ID23-2 beamline at the ESRF, Grenoble [the image was generated with the program ADXV (http://www.scripps.edu/~arvai/adxv.html); the resolution at the detector edge is 2.68 Å].

The data set was submitted to the BALBES server for automated molecular replacement at York Structural Biology Laboratory (Long et al., 2008 ▶). An initial solution with a BALBES-modified template based on a cupin-like protein (tm1010) from Thermotoga maritima (PDB entry 2f4p; Joint Center for Structural Genomics, unpublished work) contained ten copies in the asymmetric unit. It was manually inspected and the number of copies was reduced to eight owing to missing electron density for two of the ten chains. The resulting structure was then rebuilt using the PHENIX AutoBuild wizard (v.1.7.2_869; Adams et al., 2010 ▶).

Data-processing and current model statistics are presented in Table 1 ▶.

Table 1. Data-processing and initial model statistics (Winn et al., 2011 ▶; Evans, 2006 ▶).

Values in parentheses are for the highest resolution shell.

Data processing
Wavelength (Å)	0.87260
Beam dimensions (µm)	8 × 5
Space group	C2221
Unit-cell parameters (Å, °)	a = 125.29, b = 254.63, c = 82.99, α = β = γ = 90.0
Resolution limits (Å)	56.21–2.68 (2.83–2.68)
Rmerge	0.196 (0.567)
Rmeas (all I+ and I−)	0.221 (0.644)
Rp.i.m. (all I+ and I−)	0.082 (0.240)
Total No. of observations	270727 (37874)
No. of unique observations	37627 (5399)
〈I/σ(I)〉	8.4 (4.1)
Completeness (%)	100.0 (100.0)
Multiplicity	7.2 (7.0)
Wilson B factor (Å2)	40.947
Model refinement
Matthews coefficient† (Å3 Da−1)	4.96
Solvent content† (%)	75.21
Resolution range (Å)	56.2–2.7
Minimum [Fobs/σ(Fobs)]	1.34
Completeness for range (%)	99.97
No. of reflections	37602
Fit to data used in refinement
R (working + test set)	0.2283
R (working set)	0.2264
Rfree	0.2632
Free R test-set size (%)	5.03
Free R test-set count	2000

^†For eight monomers in the asymmetric unit.

3. Results and discussion  

AciX9_0562 was successfully expressed from a pET-based vector with an N-terminal TEV-cleavable His tag with high yield in the cytoplasm of E. coli BL21 (DE3) Gold strain (>50% of the total protein content as judged by SDS–PAGE). The His tag was cleaved from half of the sample. Both portions were purified by gel filtration and eluted as one symmetric peak from both Superdex 75 and Superdex 200 columns. Protein preparations with and without His tag were used to set up crystallization trials. Only protein without His tag produced crystals with sufficient quality for high-resolution structure determination. X-­ray diffraction data from a single crystal were collected to 2.7 Å resolution on the ID23-2 beamline at the ESRF, opening the way for structural studies and analysis.

Most cupins are metal-binding proteins with Mn, Fe, Ni, Cu or Zn bound in the active site. The characteristic cupin domain comprises two conserved motifs, both of which are involved in metal-ion binding. The conserved motifs comprise the following amino-acid sequences: G-X ₅-H-X-H-X _3,4-E-X ₆-G (motif 1) and G-X ₅-P-X-G-X ₂-H-X ₃-N (motif 2) (Dunwell et al., 2004 ▶). Two His residues and a Glu/Gln residue in motif 1 as well as the His residue in motif 2 act as ligands for binding of the active-site metal ion. At the current stage of refinement, a large unidentified electron density suitable for placement of the the hypothetical metal ion can be observed.

The biochemical and biophysical characterization as well as final and complete structure determination and refinement are currently in progress; the detailed results will be reported elsewhere.