Crystal Structure of Hypothetical Protein YfiH From Shigella flexneri at 2 Å Resolution

Introduction

The crystal structure of an uncharacterized conserved protein (residues 1–243) expressed by yfiH gene of Shigella flexneri 2a str.2457T (gi 30042248),¹ has been determined and refined to 2.01 Å by single wavelength anomalous dispersion (SAD) method. The YfiH protein belongs to a vast protein family of at least 201 uncharacterized proteins which contain conservative Pfam motif PF02578 (DUF152) (residues 29 –243) and TIGR motif (TIGR00726)^2,3 (residues 25–243). The YfiH protein belongs to the COG1496 which consists of 39 proteins widely distributed in 37 species from bacteria to human. The yfiH gene (DNA bases 101,672-102,403)¹ is located between the two genes, clpB and sfhB. The clpB gene is thought to encode an adenosine triphosphatase subunit of an intracellular adenosine 5′-triphosphate-dependent protease which belongs to the ClpA/ClpB family and is induced by heat shock.⁴ The sfhB gene expresses a protein which is known as suppressor of ftsH mutation and is responsible for synthesis of pseudouridine from uracil at three positions in 23S ribosomal RNA.⁵ The function of YfiH protein is unknown, but preliminary data obtained for Brevibacterium lactofermentum ATCC 13869 show that the gene is not essential for the cell’s growth and viability.⁶

Results

The seleno-methionine derivative of YfiH protein crystallized in the P1 space group with unit cell dimensions of a = 43.87, b = 50.58, c = 55.35 Å, α = 90.32°, β = 96.32°, and γ = 90.40°. There are two protein molecules in the asymmetric unit related by an approximate 2₁ symmetry along the crystallographic c axis. Both molecules adopt nearly identical backbone conformation (σ = 0.31 Å using all Cα atoms) with only a few differences in side-chain conformations. A PQS⁷ search predicted a monomeric form for this protein consistent with weak interface interaction between the two molecules in the asymmetric unit. The monomeric character was confirmed by size-exclusion chromatography (see Materials and Methods), which indicates that YfiH is a monomer in solution and this form may represent the biologically relevant species in vivo.

The protein structure is an α/β/α fold: two layers of β-sheets, the smaller one with three antiparallel strands (S1, S10, and S4), and the bigger one with six antiparallel/parallel strands (S7, S5, S6, S3, and S2) are sandwiched between two α-helices (H2, H5) on one side and one α-helix (H1) on the other [Fig. 1(A, B)]. The N-terminal end of the H1 helix is connected to S1 in the smaller β-sheet via one of the long loops and the C-terminal end of this helix is linked to S2 in the larger β-sheet. The α-helix H2, located between S6 and S7 in the larger β-sheet, contains a kink at Gly-133 and connects to S7 by an α-helical turn. In addition, there is a small domain inserted between S7 in the larger β-sheet and the helix H5. This small domain contains a small helix-turn-helix motif (H3 and H4), followed by a β-hairpin (S8 and S9). The main α/β/α fold resumes with H5, a long loop containing a little more than one 3/10-helix H5, followed by the S10 located in the middle of the smaller β-sheet.

Fig. 1.

Crystal structure of YfiH protein. A: Ribbon diagram of YfiH protein. Helices are shown in blue, β-sheets in red, and Zn^{2 +} ions are indicated in magenta; orange indicates protein residues interacting with Zn²⁺ ions, among eight Zn²⁺ ions in the asymmetric unit, six Zn²⁺ ions are shown here, four of these are shared between the two protein molecules. The two Zn^{2 +} ions not shown interact with the other protein molecule in the asymmetric unit. Secondary structure elements are also indicated in black. B: Diagram showing the secondary structure elements in YfiH protein superimposed on its primary sequence. Residues interacting with Zn²⁺ ions are marked with blue dots and residues interacting with acetate molecules are marked with red. β-Hairpins are depicted as red loops. C: The potential active site of YfiH protein. The Zn^{2 +} ions are coordinated to an acetate molecule, a cysteine and two histidine residues forming a tetrahedral configuration.

There are eight Zn⁺² ions and five acetate molecules (both were included in crystallization conditions), found in two protein molecules in the asymmetric unit (molecules A and B). Each protein molecule holds two Zn⁺² ions and the remainder of the Zn⁺² ions are shared between two protein molecules. In molecule A, one of the Zn⁺² ions is interacting with Nε2s of His-124 and His-71, Sγ of Cys-107, and an oxygen atom of one of five acetate molecules [Fig. 1(C)]. The second Zn⁺² ion coordinates Oγ1 of Thr-219, Sγ of Cys-216, and an oxygen atom from one of the acetate molecules. In molecule B, these Zn⁺² ions are involved in similar interactions. The remaining four Zn⁺² ions contribute to crystal packing. Two are shared between the two protein molecules in the same asymmetric unit and are coordinated by N ε2s of His-186 and His-39 from the two different protein molecules, and two water molecules. The other two Zn⁺² ions are coordinated by His-46, Glu-208, and Glu-147 (only for molecule A) from the two symmetry-related molecules.

In an effort to annotate the function of this protein, we searched a number of databases including BLAST,² Pro-Func,³ PQS,⁷ DALI,⁸ and ISREC-TMpred⁹ servers. Residue conservation analysis showed 201 matching sequences found by PSI-BLAST with nearly all of them being conserved hypothetical proteins from different species except for a putative inner membrane protein from Salmonella enterica Typhimurium LT2.¹⁰ The ISREC-TMpred server⁹ identified two possible transmembrane helices: 1) inside to outside helix from residues 90 (94) to 114 (112) with a score 804, and 2) outside to inside helix from residues 95 to 113 with a score of 691 (scores above 500 considered significant). However, these overlapping regions are not in any of α-helix in the YfiH structure but, nonetheless, consist of S3, S4, S5, and the connecting loops located in the two layers of β-sheets surrounded by α-helices forming a part of a hydrophobic core.

The ProFunc³ and DALI⁸ searches found several structural homologs including uncharacterized protein from Bacillus stearothermophilus¹¹ [Protein Data Bank (PDB) ID: 1t8h, Z = 11.0; sequence identity 31.0%], cytidine deaminase from Bacillus subtilis¹² (PDB ID: 1ux1, Z = 4.4; 20.8%), and cytosine deaminase from Saccharomyces cerevisiae^13,14 (PDB IDs: 1p6o, Z = 3.7; 21.2%, and 1uaq, Z = 3.6; 21.2%). The structure of conserved hypothetical protein from Salmonella enterica subsp enterica¹⁵ (PDB ID: 1rw0, sequence identity of 87%), which was not found in the DALI search because of late update of its library, is virtually identical (root-mean-square differences in Cα atoms are 0.45– 0.57 Å between chains) as expected from the sequence identity. The next closest structural homolog (1t8h) is also a conserved hypothetical protein of unknown function. Other structural homologs include cytidine and cytosine deaminases with Z scores between 4.4 –3.6. Both enzymes are the members of the pyrimidine salvage pathway, and both utilize a catalytic zinc ion and display similar mechanism. Reaction proceeds through the stereo-specific addition of a metal-bound hydroxyl group to the substrate, forming a tetrahedral transition state intermediate. Cytosine deaminase (or cytosine aminohydrolase) mediates reactions of cytosine to uracil. Cytidine deaminase correspondingly converts cytidine to uridine. These deaminases also adopt α/β/α sandwich folds with one α-helix flanked one side and two α-helices the other. However, they have only one β-sheet of four parallel strands and all α-helices are straight with no break. The structure of cytosine deaminase from S. cerevisiae (1p6o) barely superposes with the half of the large β-sheet (S7, S5, and S6) and two α-helices (H2 and H4) in the same side of the α/β/α sandwich of the YfiH structure (data not shown). In the active site of the cytidine deaminase, the Zn⁺² ion was tetrahedrally coordinated by the Nγs of Cys-291 and Cys-294, the Nδ1 of the His-262, and by the hydroxyl oxygen O4 of the reaction transition state inhibitor 4(R)-hydroxyl-3,4-dihydropyrimidine (HPY).¹³ In the YfiH protein structure for both molecules in the asymmetric unit, one of Zn⁺² ions, located in the similar area as in the structure 1p6o, is also tetrahedrally coordinated with the Nε2s of His-71 and His-124, the Sγ of Cys-107, and the OXT of an acetate molecule with the acetate molecule assuming the position occupied by HPY in the cytidine deaminase active site.¹³ Interestingly, the YfiH structure in the other crystal form, which is not included in this report, contains a similar Zn coordination, although the crystals were grown without zinc acetate. His-71, Cys-107, and His-124 are highly conserved in most of YfiH family proteins including proteins not only from archaea and bacteria but also from Homo sapiens (Q8IV20, 27% homology), filefish (CAG1329, 39%), frog (AAH78069, 31%), and mouse (Q8BZT9, 27%) (Fig. 2). However, the remainder of the residues in this potential active site are not similar to those in the structures 1p6o and of other reported cytosine/cytidine deaminases.

Fig. 2.

Multiple sequence alignment of YfiH protein homologs from various species. AAP17973.1 (Shigella flexneri), NP_754996.1 (Escherichia coli), NP_461592.1 (Salmonella typhimurium), AB0832 (Salmonella enterica subsp enterica), ZP_00156016.2 (Haemophilis influenzae), CAG13296.1 (Tetraodon nigroviridis), AAH78069.1 (Xenopus laevis), XP_233749.1 (Ratus norwegicus), AAH35749.1 (Homo sapiens), XP_417036.1 (Gallus gallus), NP_766076.1 (Mus musculus). Idt, percentage of identity. The conservations of His-71, Cys-107, and His-124, and identical amino acids are shown in red and blue, respectively.

Because there are limited functional data (other than structural studies) of cytosine and cytidine deaminases, it is difficult to determine if the two families are functionally related. Because the sfhB, the pseudouridine synthase gene, and yfiH gene with possible cytosine/cytidine deaminase function are neighbors, it is conceivable to say that YfiH protein could participate in the same pyrimidine salvage pathway as sfhB gene product.

Materials and Methods

Protein cloning, expression, and purification

The YfiH gene was cloned in pMCSG7 vector¹⁶ and overexpressed in Escherichia coli BL21 (DE3)—Gold (Stratagene) harboring an extra plasmid encoding three rare tRNAs (AGG and AGA for Arg, ATA for Ile). The pMCSG7 vector bearing a TEV protease cleavage site creates a construct with cleavable His₆-tag fused into N-terminus of the target protein and adds three artificial residues (SerAsnAla) on that end. The cells were grown using SeMet-containing enriched M9 medium and conditions known to inhibit methionine biosynthesis.^17,18 The cells were grown at 37°C to an OD₆₀₀ of ~0.6 and protein expression induced with 1 mM isopropyl-β-D-thiogalactopyransoid. After induction, the cells were incubated overnight with shaking at 20°C. The harvested cells were resuspended in 5 volumes of lysis buffer (50 mM HEPES pH 8.0, 500 mM NaCl, 10 mM imidazole, 10 mM β-mercaptoethanol, and 5% v/v glycerol) and stored at −20°C.

The thawed cells were lysed by sonication after the addition of inhibitors of proteases (Sigma, P8849) and 1 mg/mL lysozyme. The lysate was clarified by centrifugation at 30,000g (RC5C-Plus centrifuge, Sorval) for 20 min, followed by filtration through a 0.45-μm filter and 0.22-μm in-line (Gelman).

The standard purification protocol was thoroughly described previously.¹⁹ Immobilized metal affinity chromatography (IMAC-I) using a 5-mL HiTrap Chelating HP column charged with Ni⁺² ions and buffer-exchange chromatography on a HiPrep 26/10 desalting column (both Amersham Biosciences) were performed using AKTA EX-PLORE 3D (Amersham Biosciences). His₆-tag was cleaved using the recombinant TEV protease expressed from the vector pRK508.²⁰ The protease was added to the target protein in a 1:30 ratio and the mixture was incubated at 4°C for 48 h. The YfiH protein was then purified using a 1-mL HiTrap Chelating column (Amersham Biosciences) charged with Ni⁺² ions. The protein was dialyzed in 20 mM Tris-HCl pH 7.1, 50 mM NaCl, 2 mM DTT, and concentrated using a Centricon Plus-20 Centrifugal Concentrator (Millipore).

Protein assays

The molecular weight of YfiH protein in solution was determined by size-exclusion chromatography on Superdex-75 10/30 column (Amersham Biosciences) calibrated by ribonuclease (13.7 kDa), chymotrypsinogen (25 kDa), ovalbumine (43 kDa), albumin (67 kDa), as standards. The calibration curve of K_av = V_e − V_o/V_t − V_o was used, where V_e is the elution volume for the protein, V_o is the column void volume, and V_t is the total bed volume.

Protein crystallization

The protein crystals initially were obtained by using crystallization screen Index (Hampton Research) and Wizard I and II crystallization screens (Emerald Biostructures) with Cartesian robot system (Cartesian Technologies) using sitting drop technique. The crystals appeared in 1 day in five conditions: Index #4 and #93 and Wizard II #4, #11, and #42. The protein was then crystallized by optimization of the Index condition #93 by vapor diffusion in hanging drops by mixing 2 μL of the protein solution (30 mg/mL) with 1 μL of 0.1 M zinc acetate and 14% (w/v) polyethylene glycol 3350, and equilibrated at 23°C over 1 mL of this solution. Crystals, which appeared the next day, were flash-frozen in liquid nitrogen with crystallization solution containing 15% (v/v) glycerol as cryoprotectant before data collection.

Data collection

Diffraction data were collected at 100° K at the 19ID beam line of the Structural Biology Center at the Advanced Photon Source, Argonne National Laboratory. The SAD data at 0.9793 Å (peak energy: 12.6603 KeV) near the Se absorption edge up to 2.01 Å were collected from a single (0.1 × 0.02 × 0.05 mm) Se-Met labeled protein crystal in the cold stream maintained around 100° K. The crystal was exposed for 8 s per 1.5° rotation of ω with the crystal to detector distance of 200 mm. The data were recorded on an SBC-2 detector by the two scannings of 225° on ω, at ϕ = 0° and ϕ = 180°. The space group was P1 with cell dimension of a = 43.87, b = 50.58, c = 55.35 Å, α = 90.32°, β = 96.32°, and γ = 90.40°. All data were processed and scaled with HKL2000 suite²¹ (Table I) to an Rmerge of 7.1%.

TABLE I.

Summary of the YfiH Crystal Data and SAD Data Collection

Unit cell parameters	a = 43.874, b = 50.583, c = 55.347 Å
	α = 90.32°, β = 96.32°, γ = 90.40°
Space group	PI
Molecular weight [243 residues (SeMet)]	26,380 Da
Molecules per asymmetric unit (a.u.)	2
SeMet residues per a.u.	8
SAD data collection
Wavelength (Å)/energy (keV)	0.9793/12.6603
Resolution range (Å)	30–2.01 (2.08–2.01)a
No. of unique reflections	30,856 (2,849)
Completeness (%)	97.4 (89.6)
Rmergeb	0.071 (0.21)

^aIn parentheses are statistics for the highest resolution shell.

^bR merge = (|Ihkl − 〈I〉|)/Ihkl, where the average intensity 〈I〉 is taken over all symmetry equivalent measurements and Ihkl is the measured intensity for any given reflection.

Structure determination and refinement

The structure was determined by SAD phasing utilizing the anomalous signal from Se atoms using HKL2000_PH (W. Minor, University of Virginia, personal communication) and RESOLVE²² and refined to 2.0 Å using REFMAC 5.2²³ in CCP4.²⁴ The initial model was completed by using ARP/wARP²⁵ and manual tailoring using Coot.²⁶ The structure 1rw0¹⁵ was used as a guide and to confirm the models. The final R was 0.18 with the free R of 0.25 with all data (Table II). Electron density calculated at 1.2 σ is well connected for most of the main-chain except a few areas on the surface of the molecules.

TABLE II.

Refinement Statistics for YfiH Structure

Resolution range (Å)	55.05–2.01
No. of unique reflections	27,582
Sigma cutoff	0.0
R-valuea	0.179
R-freeb	0.252 (3071)c
RMS deviations from ideal geometry
Bond length (1–2) (Å)	0.015
Angle (°)	1.517
No. of atoms
Protein	4,148
Acetate	20
Glycerol	12
Zn	8
Water	420
Mean B-factor (Å2)
all atoms	22.5
Protein atoms
Protein main-chain	20.5
Protein side-chain	21.8
Acetate	21.6
Glycerol	19.4
Zn	29.0
Water	33.2
Ramachandran plot statistics (%)
Residues in allowed regions	99.0 (479)
Residues in disallowed region	1.0 (4)

^aR-value = ||F_o| − |Fc||/||F_o|, where F_o and F_c are the observed and calculated structure factor amplitudes, respectively.

^bR-free is equivalent to R value but is calculated for 10% of the reflections (number in the parentheses).

^cChosen at random and omitted from the refinement process.

Validation and deposition

The stereochemistry of the structure was checked with PROCHECK²⁷ and the Ramachandran plot. The main-chain torsion angles for all residues except four are in allowed regions. The main-chain conformations of these four residues (Asp-106 and Lys-231 per molecule), although, appear not in allowed regions, are well supported by the electron density (2fo-fc map contoured at 1.5 σ) suggesting that they are also in proper conformation. Atomic coordinates and experimental structure factors of YfiH have been deposited with the PDB and are accessible under the code 1xaf.