Evaluation of Thermal Hysteresis Activity of Ice-binding Proteins Using Ice-etching and Molecular Docking

The recombinant FfIBP was expressed and purified as described previously.³⁵ Briefly, the E. coli BL21 strain, harboring the FfIBP gene in the pColdI vector, was grown at 37 °C in Luria-Bertani (LB) medium containing 100 µg/ml ampicillin with shaking until the OD₆₀₀ reached 0.6. Then, the bacterial culture was transferred to 15 °C and further incubated for 1 h. The recombinant FfIBP was induced by adding isopropyl beta-D-thiogalactopyranoside to a final concentration of 1 mM, and the cells were further incubated at 15 °C for 20 h. The harvested cell pellet was resuspended and incubated for 30 min in 50 mM sodium phosphate pH 8.5, 300 mM NaCl, 5 mM imidazole, 0.5 mg/ml lysozyme, and 1 mM phenylmethylsulfonyl fluoride. The cell suspension was disrupted by ultrasonication (Vibra-Cell VCX400) and centrifuged at 12,000 rpm for 1 h at 4 °C. The supernatant was loaded onto a His-tag affinity column. The column was washed with ten column volumes of lysis buffer and eluted with 50 mM sodium phosphate pH 8.5, 300 mM NaCl, 300 mM imidazole. The eluent, containing FfIBP, was dialyzed against the lysis buffer. The concentration of FfIBP was determined using UV spectrophotometry with an extinction coefficient at 280 nm of 26930 M⁻¹ cm⁻¹.

Single ice crystal preparation and the ice-etching experiment were performed as described elsewhere with slight modification.¹⁴ A single ice crystal (2.5 cm × 2.5 cm × 2.5 cm) was mounted to the cold-finger to the long axis, of which either the basal or primary prism planes of the ice crystal were oriented. The mounted ice crystal was placed into degassed water (4 °C), and the temperature of the cold-finger was lowered and maintained at −7 °C for 3 h by circulating a coolant. In this stage, the ice crystal was grown into the ice hemisphere. Then, the ice hemisphere was dipped into the IBP-containing solution (1~5 µM), and the cold-finger was cooled and maintained for 30 min at −10 °C. The FfIBP-bound ice hemisphere was detached from the cold-finger and transferred to a cold room at −10 °C. To expose the distinct etching pattern, the surface of the ice hemisphere was scraped briefly using a sharp blade and left to evaporate for at least 1 h.

LeIBP (PDB ID: 3UYV) and FfIBP (PDB ID: 4NU2) structural files were obtained from RCSB PDB (www.rcsb.org). Four types of ice crystal planes, i.e.: the primary prism {1 -1 0 0}, secondary prism {2 -1 -1 0}, pyramidal {1 -1 0 1}, and basal {0 0 0 1} planes, were generated by VESTA software, version 3.2.1.³⁶ The interaction poses between IBPs and those ice crystals were searched using the Hex 8.0 docking program.³⁷ Default parameters of Hex 8.0 were used, except for the correlation type (Shape + Electrostatic), post-processing (OPLS Minimisation), and final search (30). The best predicted conformations were saved, and the shape complementarities (Sc) were calculated using the SC program in the CCP4 program suite.³⁸

A single ice hemisphere, whose either basal plane or primary prism plane was oriented normal to the finger, was immersed in a FfIBP-containing solution, as shown in Fig. 3. There was no unique ice-etching pattern formed by FfIBP, regardless of the ice crystal orientation (Fig. 3A and B), indicating that it binds to the entire surface of the ice hemisphere. A handful of data showed that absorption of IBPs onto the multiple ice planes seems to be mediated by their compound IBSs.^31,33,39 Mutagenesis and GFP-tagging of moderately active fish type III AFP variant QAE revealed it binds to both primary and pyramidal ice planes by two adjacent and flat IBSs forming an angle of 150° from each other. primary, and pyramidal ice plane binding sites forming an angle of 150° from each other, each binding primary and pyramidal ice planes.³¹ However, moderately active fish type I AFPs are shown to bind onto only single ice planes: AFPs from winter flounder (Pseudopleuronectes americanus) and Alaskan plaice (Pleuronectes quadritaberulatus) bound to the pyramidal planes of ice, while the Sculpin (Myoxocephalus Sc orpius) AFP to the secondary prism planes.¹⁴ Additionally, hyperactive TisAFP and ColAFP, two closest homologs of FfIBP and LeIBP, also bound to multiple ice planes by two IBSs on the flat β-sheet and the adjacent loop.^33,39 Based on the ice-etching experiments, both LeIBP and FfIBP bound to multiple ice planes, however, it is not clear how many IBSs exist in FfIBP and LeIBP due to the limited site-directed mutagenesis study.

Figure3.

Ice etching photographs of FfIBP. Top view (upper panel) of a single ice crystal hemisphere grown in a FfIBP-containing solution with its c-axis parallel (A) or normal (B) to the cold-finger. (Lower panel) Schematic drawing (lower panel) of each hemisphere grown with its c-axis parallel (A) or normal to (B) the cold-finger. In the right panel, the shaded areas are the basal and primary prism planes.

As other hyperactive IBPs, for example ColAFP, hyperactive FfIBP recognized multiple ice planes, such as the basal, primary prism, and secondary prism planes.^11,33,39-44 However, previously we have also shown that even moderately active LeIBP binds, relatively weakly, to the basal and primary prism planes.¹⁷ Since the binding onto the basal plane was hypothesized to be the hallmark of hyperactivity of IBPs,16,40 the ice-etching observation per se may demonstrate that LeIBP may not be moderately active, but hyperactive.¹⁷ On the contrary to this hypothesis, our result indicated that the binding onto the basal plane cannot guarantee the hyperactivity of IBPs. Hence we employed the docking and shape complementarity to further evaluate the strength of the interaction between the IBPs and specific ice planes.

The flatness of the ice binding site (IBS) is considered important for the IBP’s function, which provides an ideal geometrical fit to bind ice planes.²² As shown in Fig. 2, LeIBP and FfIBP roughly have 3 faces (A, B, and C faces), which are relatively flat. Mutagenesis revealed that the IBSs of LeIBP and FfIBP are on the B faces.¹⁹

In order to gain better insight into the interactions of LeIBP and FfIBP with ice, docking was performed to acquire the lowest energy interactions between the IBPs and the ice crystals using Hex 8.0 docking software, with the parameters listed in Materials and Methods. The B-face of each IBP was docked onto the four ice crystal planes: the primary prism, secondary prism, pyramidal, and basal planes. The shape complementarity (Sc) of the lowest energy interactions between IBPs and the ice crystals were then calculated to evaluate the interaction of the IBP and ice planes (Table 1). A perfect pair of two surfaces has a Sc value of 1.0. Generally, IBPs have a Sc value between 0.5 and 0.8.^25,33,39 LeIBP and FfIBP interactions with multiple ice planes are shown in the Figs. 4 and 5.

Figure4.

Docking of LeIBP to different ice planes (red spheres are water molecules that contribute directly to binding to the ice crystal). (A) basal, (B) pyramidal (C) primary prism, and (D) secondary prism planes.

Figure5.

Docking of FfIBP to different ice planes (red spheres are water molecules that contribute directly to binding to the ice crystal). (A) basal, (B) pyramidal (C) primary prism, and (D) secondary prism planes.

Lysozyme (PDB ID: 3A3Q) was used as a control and was docked onto the {1 -1 0 0} primary prism ice plane with a lower Sc value (only 0.27). This value agrees with previous results (<0.38).^25,33 This result clearly indicates that the Sc value can be used to evaluate the interaction between IBPs and ice. Garnham et al. (2011) showed that bound water molecules on the IBS are directly involved in binding to ice. It was hypothesized that the bound waters relatively regularly embedded in the IBS enable IBPs to bind to ice planes.^24,44 Hence, we removed the water molecules and recalculated the Sc values. As shown in Table 1, the Sc values of IBPs with bound water molecules were higher than those of IBPs without bound water. The result is consistent with prior reports, showing that bound water molecules of a similar arrangement to the ice lattice play an important role in forming a H-bond network with ice.^24,33,39 The present docking study verified that water molecules bound to the IBS of IBPs clearly increase the degree of fitness between IBPs and the specific ice planes.

The calculated Sc values of FfIBP in the present study were higher in all cases than those of LeIBP, implying that the IBS of FfIBP has better complementarity to the four ice planes than that of LeIBP. The contact area of the IBS with the primary prism, pyramidal, and basal planes was similar in both IBPs, but that of the IBS with the secondary prism plane was larger in FfIBP (1072.68 Ų), than in LeIBP (809.39 Ų). Some of these interactions are displayed in Open book view in Fig. 6. A previous report also showed that hyperactive TisAFP8 also had higher Sc values than its moderately active counterpart, TisAFP6.³⁹ The loop region of the hyperactive TisAFP8 had an Sc value of 0.55 with contact area of 809.39 Ų while that of moderately active TisAFP6 had an of 0.49 with contact area of 783 Ų. The Sc value and contact area were different structurally similar two IBP cases. These differences may lead to the difference in TH activities. Taken together, these data may explain why FfIBP is hyperactive compared to LeIBP.

Figure6.

Open book view of interaction between FfIBP and pyramidal plane (A), and between LeIBP and secondary prism plane (B).

The Sc values can also partly support the observations in the ice-etching experiment. Previously, LeIBP adsorbed preferentially onto the basal, and relatively weakly to the primary prism planes. The docking results indicated that LeIBP with bound water molecules within the IBS additionally binds to the secondary prism plane with a slightly higher Sc value than to the basal and primary prism planes (Table 1). However, the contact area between the IBS of LeIBP and the secondary prism plane was quite smaller (809.39 Ų) than those between LeIBP and the primary prism (1093 Ų) and basal (1195 Ų) planes. The smaller contact area may account for the no observation of etching pattern of the secondary prism plane in LeIBP, which is consistent with the observation that larger contact surface areas probably compensate for the relatively low Sc values.³³ In addition, the interaction between LeIBP and the pyramidal plane was negligible, based on the Sc value of 0.39, which is slightly above the values for the non-IBP interaction. The Sc value of FfIBP indicates that it binds to all four planes with almost the same values and contact areas, which agrees with the result of the ice-etching experiment. Similarly, the structurally highly similar ColAFP also bound to all four planes with almost the same values.³³ Based on the present observations, the hyperactivity of IBP may be explained by the shape complementarity and contact area between IBS of IBPs and ice crystal, which is intensified by the bound water molecules within IBS. However, the further investigations are needed to answer the difference in TH activity between structurally similar IBPs.