Crystallization and preliminary X-ray crystallographic analysis of an ice-binding protein (FfIBP) from Flavobacterium frigoris PS1

The ice-binding protein FfIBP from F. frigoris PS1 was overexpressed, purified, characterized and crystallized, and preliminary X-ray crystallographic analysis was performed.

Abstract

Ice growth in a cold environment is fatal for polar organisms, not only because of the physical destruction of inner cell organelles but also because of the resulting chemical damage owing to processes such as osmotic shock. The properties of ice-binding proteins (IBPs), which include antifreeze proteins (AFPs), have been characterized and IBPs exhibit the ability to inhibit ice growth by binding to specific ice planes and lowering the freezing point. An ice-binding protein (FfIBP) from the Gram-negative bacterium Flavobacterium frigoris PS1, which was isolated from the Antarctic, has recently been overexpressed. Interestingly, the thermal hysteresis activity of FfIBP was approximately 2.5 K at 50 µM, which is ten times higher than that of the moderately active IBP from Arctic yeast (LeIBP). Although FfIBP closely resembles LeIBP in its amino-acid sequence, the antifreeze activity of FfIBP appears to be much greater than that of LeIBP. In an effort to understand the reason for this difference, an attempt was made to solve the crystal structure of FfIBP. Here, the crystallization and X-­ray diffraction data of FfIBP are reported. FfIBP was crystallized using the hanging-drop vapour-diffusion method with 0.1 M sodium acetate pH 4.4 and 3 M sodium chloride as precipitant. A complete diffraction data set was collected to a resolution of 2.9 Å. The crystal belonged to space group P4₁22, with unit-cell parameters a = b = 69.4, c = 178.2 Å. The asymmetric unit contained one monomer.

1. Introduction  

Many plants, insects, animals and other organisms have evolved unique adaptive mechanisms that allow them to survive in harsh environments with extreme temperatures (Rothschild & Mancinelli, 2001 ▶). To survive in extremely cold environments, many organisms produce antifreeze proteins (AFPs) and ice-binding proteins (IBPs). These proteins can lower the freezing point of a solution non­colligatively without affecting the melting point of the solution (thermal hysteresis); thus, the freezing of body fluids can be prevented. In addition, these proteins can inhibit ice recrystallization, where large ice crystals grow at the expense of smaller ones, thus preventing cell damage during freeze–thaw cycles (Knight et al., 1984 ▶, 1991 ▶). It is generally accepted that these proteins function through adsorption of their flat ice-binding surfaces onto particular planes of ice crystals, thus preventing or inhibiting further ice growth (Yeh & Feeney, 1996 ▶). Notably, the antifreeze activities of AFPs and IBPs vary and depend on the protein structure. Therefore, structural studies are essential to understand the antifreeze and ice-binding mechanisms of these proteins.

AFPs are characterized according to their structure and their thermal hysteresis (TH) values and have been categorized into several types: types I–IV from fish, insect AFPs, bacterial AFPs and plant AFPs (Griffith et al., 1997 ▶; Davies & Hew, 1990 ▶; Graether et al., 2000 ▶). Recently identified IBPs from bacteria, diatoms and fungi form a distinct cluster according to phylogenetic analysis (Raymond & Janech, 2009 ▶). The first ice-active fungal protein was found in a snow mould, Typhula ishikariensis (Hoshino et al., 2003 ▶), and similar ice-binding proteins were subsequently found in sea-ice diatoms (Janech et al., 2006 ▶), a sea-ice bacterium (Raymond et al., 2007 ▶), a bacterium from a deep ice core in the Antarctic ice sheet (Raymond et al., 2008 ▶) and the Arctic yeast Leucosporidium sp. AY30 (Lee et al., 2010 ▶). Recently, the crystal structure of IBP (LeIBP) from Leucosporidium sp. AY30 was determined in our laboratory. Structural and functional analyses of LeIBP have shown that this protein has a β-helical fold similar to those of insect and bacterial AFPs. However, the ice-binding site of LeIBP is more complex and does not have a simple ice-binding motif. Insect and bacterial AFPs use a Thr-X-­Thr motif and a Thr-X-Asx motif, respectively, located in a flat β-­sheet to bind ice (Lee et al., 2012 ▶).

In this study, we succeeded in expressing and purifying an IBP (FfIBP) from an Antarctic bacterium, Flavobacterium frigoris PS1, isolated from sea ice on the shore of McMurdo Sound. The ice-binding property of FfIBP has previously been confirmed by ice-pitting experiments (Raymond et al., 1989 ▶). FfIBP shows 56% amino-acid sequence similarity and 39% identity to LeIBP. However, FfIBP has up to a tenfold higher antifreeze activity compared with LeIBP. To determine the reason for this difference in activity, we performed structural and functional studies. We report the results of preliminary X-ray crystallographic experiments on FfIBP, which are the first step towards structural elucidation of FfIBP.

2. Materials and methods  

2.1. Cloning, expression and purification of recombinant FfIBP  

The FfIBP gene (amino acids 29–276), excluding the signal peptide sequence (amino acids 1–23) identified by SignalP (Petersen et al., 2011 ▶), was amplified by polymerase chain reaction (PCR) using F. frigoris PS1 genomic DNA as a template. The forward primer was 5′-CGA TAA CAT ATG TCT CTA TCA GTT GCA AAT-3′ and the reverse primer was 5′-CGA TAA CTC GAG TCA TTG TGG TAT GGT AAC GGT-3′. Amplified fragments were cleaved with NdeI and XhoI and cloned into the pCold I vector (Takara Bio Inc., Japan) to produce a factor Xa protease-cleavable N-terminally 6×His-tagged fusion protein. The construct was verified by DNA sequencing and transformed into Escherichia coli strain BL21 (DE3) for expression. The cells were grown in LB medium at 310 K in the presence of 100 µg ml⁻¹ ampicillin until the OD₆₀₀ approached 0.6. Recombinant protein expression was then induced using 1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) at 289 K for 24 h. The cell pellet was harvested by centrifugation and resuspended in cold lysis buffer consisting of 50 mM sodium phosphate pH 8.5, 300 mM NaCl, 5 mM imidazole, 0.5 mg ml⁻¹ lysozyme, 1 mM phenylmethylsulfonyl fluoride (PMSF). After 30 min incubation at 277 K, the cells were disrupted by ultrasonication (Vibra-Cell VCX400). The lysate was centrifuged again at 16 000 rev min⁻¹ for 40 min at 277 K to remove cell debris and the supernatant was poured into a His-tag affinity column. The bound FfIBP was washed with ten column volumes of lysis buffer and eluted with elution buffer consisting of 50 mM sodium phosphate pH 8.5, 300 mM NaCl, 300 mM imidazole. After cleavage of the 6×His tag at 277 K overnight, size-exclusion chromatography was performed as the next step of purification. The sample was applied onto a Superdex 200 HiLoad 16/60 column equilibrated with 20 mM Tris–HCl pH 8.5, 150 mM NaCl.

2.2. Antifreeze activity (thermal hysteresis) assay  

Measurement of the thermal hysteresis (TH) activity was performed as described previously (Lee et al., 2012 ▶). Briefly, the sample was placed on the surface of a temperature-controlled metal block and cooled rapidly to around 253 K using a nanolitre osmo­meter (Otago Osmometers, Dunedin, New Zealand) equipped with a Canon digital camera (PowerShot A620). After 5 min incubation, the temperature was raised slowly to obtain a single ice crystal. When only one ice crystal remained, the temperature was slowly lowered again (−0.5 K min⁻¹). The freezing point of the sample was taken as the temperature at which there was rapid ice-crystal growth. The TH value was defined as the maximum difference between the melting and freezing temperatures of the sample. The ice-crystal morphology produced by IBPs in distilled water was examined.

2.3. Crystallization and data collection  

Purified FfIBP was concentrated to 16.4 mg ml⁻¹ and crystallized by the hanging-drop vapour-diffusion method at 293 K. The drops consisted of 1 µl protein solution (16.4 mg ml⁻¹ protein in 20 mM Tris–HCl pH 8.5, 150 mM NaCl) and 1 µl reservoir solution (0.1 M sodium acetate pH 4.4, 3 M sodium chloride). A single crystal formed after 2 d and grew to a largest dimension of 0.5 mm. Successful flash-cooling was achieved when the FfIBP crystals were transferred directly into Paratone oil and allowed to equilibrate for 2 min. Native data for FfIBP were collected to a resolution of 2.9 Å using the X-ray facility home source at the Korea Basic Science Institute (KBSI), Ochang, Republic of Korea. The data were processed using iMOSFLM and SCALA (Battye et al., 2011 ▶).

The crystals belonged to space group P4₁22, with unit-cell parameters a = b = 69.4, c = 178.2 Å. The phases were determined by molecular replacement using MOLREP with the coordinates of the LeIBP structure (PDB entry 3uyu; Lee et al., 2012 ▶). After computational refinement using REFMAC5 (Murshudov et al., 2011 ▶), an interpretable electron-density map was calculated in CCP4 (Winn et al., 2011 ▶). Model building using Coot (Emsley & Cowtan, 2004 ▶) and further refinement is under way.

3. Results and discussion  

An ice-binding protein (FfIBP) from F. frigoris PS1 was cloned, overexpressed and purified in E. coli BL21 (DE3) cells (Figs. 1 ▶ a and 1 ▶ b). Purified FfIBP appeared to be a monomer as suggested by the size-exclusion chromatography elution profile and the results of analytic ultracentrifugation (data not shown). In the LeIBP structure a C-terminal hydrophobic loop adopts an extended conformation and plays in a role in dimer formation (Lee et al., 2012 ▶). However, this C-­terminal region is not present in FfIBP, which may cause FfIBP to behave as a monomer in solution. Notably, C-terminally truncated LeIBP had a greater TH activity than the wild-type protein. We had predicted that removal of the C-terminus would prevent the interactions necessary for dimerization, allowing the protein to be freely accessible for ice binding in the monomeric state. The monomeric state may confer higher TH activity on the C-terminally truncated LeIBP mutant. Most strikingly, the recombinant FfIBP displayed a TH activity of 2.5 K at 50 µM, which is ten times greater than the activity of the moderately active LeIBP (0.17 K at 50 µM; Fig. 1 ▶ c; Lee et al., 2012 ▶). Although FfIBP shares protein sequence similarity with LeIBP and is expected to have a similar overall three-dimensional structure, it exhibits distinct differences from LeIBP, including differences in oligomerization state and TH activity.

Figure 1.

Multiple sequence alignment and antifreeze activity of recombinant FfIBP. (a) Sequence alignment of FfIBP and other IBPs from Colwellia psychrerythraea (ColIBP), Leucosporidium sp. AY30 (LeIBP), Flammulina populicola (FpoIBP), Lentinula edodes (LedIBP), Typhula ishikariensis (TisIBP), Cytophaga hutchinsonii ATCC 33406 (ChuIBP), Navicula glaciei (NglIBP) and Stephos longipes (SloIBP) using ClustalX. The residues constituting the ice-binding sites in the LeIBP structure are indicated by filled circles. Highly conserved residues are shaded grey and black. (b) An aliquot of purified FfIBP was visualized by 12% SDS–PAGE. (c) Comparison of the thermal hysteresis activities of FfIBP and LeIBP. The thermal hysteresis activity of FfIBP (circles) observed was 2.5 K at 50 µM, which was approximately ten times greater than the activity of LeIBP (triangles).

To obtain structural information about the mechanisms responsible for the hyper-antifreeze activity and the differences in the ice-binding mechanism of FfIBP, we performed X-ray crystallography experiments. Purified FfIBP was concentrated to 16.4 mg ml⁻¹ and quadrangular pyramid-shaped crystals suitable for diffraction (maximum dimension of 500 µm; Fig. 2 ▶ a) were grown in 0.1 M sodium acetate pH 4.4, 3 M sodium chloride. Complete diffraction data were collected to a resolution of 2.9 Å (Fig. 2 ▶ b). The crystal belonged to space group P4₁22, with unit-cell parameters a = b = 69.4, c = 178.2 Å (Table 1 ▶). The Matthews coefficient (V _M) of 4.21 ų Da⁻¹ suggests the presence of one molecule in the asymmetric unit, with a solvent content of 70.8% (Matthews, 1968 ▶). A good molecular-replacement solution (translation-function contrast value of 18.8 and R factor of 48%) was obtained using the monomeric model of the LeIBP structure (PDB entry 3uyu) with the MOLREP software. The resulting PDB file was refined against the original data set in REFMAC5 (Murshudov et al., 2011 ▶) and had reasonable R _work and R _free values of 25.5% and 32.3%, respectively. Further structure refinement and model building of FfIBP are currently in progress; the structural details will be reported elsewhere.

Figure 2.

Crystal morphology and diffraction pattern of recombinant FfIBP. (a) FfIBP crystals grown in 0.1 M sodium acetate pH 4.4, 3 M sodium chloride. (b) X-ray diffraction image (2.9 Å resolution) obtained during data collection from the FfIBP crystal.

Table 1. Data-collection statistics.

Values in parentheses are for the highest resolution shell.

X-ray source	Rigaku MicroMax-007 HF at KBSI
Space group	P4122
Unit-cell parameters (Å)	a = 69.4, b = 69.4, c = 178.2
Wavelength (Å)	1.5418
Resolution range (Å)	69.36–2.90 (3.06–2.90)
No. of observed reflections	74340 (9447)
No. of unique reflections	10314 (1458)
Completeness (%)	99.9 (100)
Multiplicity	7.2 (6.5)
R merge †	0.147 (0.405)
〈I/σ(I)〉	9.0 (3.0)

^† R _merge = , where I _i(hkl) is the intensity of the ith observation of reflection hkl, is the sum over all reflections and is the sum over i measurements of reflection hkl.