ARTICLE pubs.acs.org/bc
Conjugation of Type I Antifreeze Protein to Polyallylamine Increases Thermal Hysteresis Activity € Ozge Can† and Nolan B. Holland* Department of Chemical & Biomedical Engineering, Cleveland State University, 2121 Euclid Avenue, Cleveland, Ohio 44115, United States ABSTRACT: Antifreeze proteins (AFPs) are ice binding proteins found in some plants, insects, and Antarctic fish allowing them to survive at subzero temperatures by inhibiting ice crystal growth. The interaction of AFPs with ice crystals results in a difference between the freezing and melting temperatures, termed thermal hysteresis, which is the most common measure of AFP activity. Creating antifreeze protein constructs that reduce the concentration of protein needed to observe thermal hysteresis activities would be beneficial for diverse applications including cold storage of cells or tissues, ice slurries used in refrigeration systems, and food storage. We demonstrate that conjugating multiple type I AFPs to a polyallylamine chain increases thermal hysteresis activity compared to the original protein. The reaction product is approximately twice as active when compared to the same concentration of free proteins, yielding 0.5 °C thermal hysteresis activity at 0.3 mM protein concentration. More impressively, the amount of protein required to achieve a thermal hysteresis of 0.3 °C is about 100 times lower when conjugated to the polymer (3 μM) compared to free protein (300 μM). Ice crystal morphologies observed in the presence of the reaction product are comparable to those of the protein used in the conjugation reaction.
’ INTRODUCTION Antifreeze proteins (AFPs) help organisms to survive at subzero temperatures or at temperatures lower than the freezing point of their body fluids by inhibiting ice crystal growth. Interest in this phenomenon has given rise to many studies in order to gain more information as to how they prevent the ice crystal growth.1 It is generally accepted that AFPs bind to specific planes of the ice surface, which stops subsequent crystal growth at temperatures below the colligative melting point. This interaction of AFPs with ice crystals results in thermal hysteresis, which is the difference between the freezing and melting temperatures. AFPs also prevent the redistribution of the water between ice crystals and the solution referred to as ice recrystallization.2 These properties have use in potential applications such as cold storage of cells or tissues, ice slurries used for refrigeration systems, and food storage.3 8 To minimize the amount of protein needed for a given application, antifreeze proteins with thermal hysteresis activity at low concentrations are desirable. Among AFPs, there is a considerable variance in thermal hysteresis activity, even among proteins with similar structures. The most extensively characterized type I fish AFP is the winter flounder HPLC6 protein, which has three repeats of 11 amino acid residues.9,10 A longer type I AFP is AFP9 from winter flounder, which has four repeat sequences.11 This longer protein is more active than its shorter isoform, illustrating the general effect of increased size on the thermal hysteresis activity. More recently, a much larger type I AFP was discovered in American plaice with a molecular weight of approximately 17 kDa. This protein is five times larger than the AFP9 and exhibited significant thermal hysteresis activity even at low concentrations.12,13 r 2011 American Chemical Society
Differences in thermal hysteresis activity have also been observed in other types of AFPs. For example, the activity of β-helical insect AFPs is affected by the addition or deletion of coils from the structure of the protein. It has been shown that enhanced activity in longer insect AFPs results from the larger ice surface binding area of the proteins, as long as the repeating sequences continue to match the ice crystal lattice structure.14,15 For the globular type III antifreeze proteins, a different mechanism for increased activity has been reported. Miura et al. observed enhanced activity in a natural protein consisting of two type III AFP domains connected by a nine residue linker.16 They compared the activity of this protein with its single domain counterpart and reported that the two-domain protein is more active than the single domain protein. This enhancement is most pronounced at low concentrations, yielding thermal hysteresis activities as much as 6 times greater than the monomer. This effect was explained by the assumption that the average distance between the two domains of the protein adsorbed on the surface is smaller than that of the adsorbed monomers. Hence, in order to saturate the ice surface and stop the ice growth, a smaller amount of dimer would be needed as compared to the monomer. A more detailed investigation of the increased activity of the type III dimers showed that a recombinant two-domain type III AFP from Antarctic eel pout is twice as active as its monomer.17 However, when they inactivated one of the domains, its activity was reduced to only 1.2 times greater than that of the monomer. Received: August 9, 2011 Revised: September 10, 2011 Published: September 12, 2011 2166
dx.doi.org/10.1021/bc2004318 | Bioconjugate Chem. 2011, 22, 2166–2171
Bioconjugate Chemistry When two monomers were linked through a disulfide bond such that two domains could not bind to the ice surface simultaneously, the dimer was again only 1.2 times as active as its monomer, indicating that the two domains need to be both active and able to simultaneously bind to the ice surface to achieve maximum activity. Expanding on this principle, thermal hysteresis activity of type III was further increased by adding domains by tandem repetition to create three and four domain proteins.18 However, diminishing return above three domains was observed. On the basis of the reported flexibility of the linker,19,20 this likely is the result of limited freedom of the domains in the multimer, particularly the central domains that are confined at both termini. To overcome the limited mobility of tandem repeats, we present here a different strategy for increasing the number of domains per molecule. Type I AFPs are conjugated to a polymer chain through their carboxy-terminal end. This results in multiple binding domains per chain, while still retaining individual domain mobility. Type I AFP, as opposed to a more active AFP, was chosen for two reasons. First of all, it consists of a single α helix with no disulfide bonds or post-translational modification, so large quantities of active protein (∼100 mg/L of culture) can be expressed in E. coli and purified using relatively simple procedures.21 Second, it could be modified to remove acidic residues that would compete with the carboxy-terminus in the conjugation reaction. The goal of the work is to reduce the amount of protein necessary to observe significant thermal hysteresis activity, so the low overall thermal hysteresis of the Type I AFP is not a primary issue.
’ MATERIALS AND METHODS Synthesis and Cloning of the Genes for the Proteins. E. coli strain BL21*(DE3) (Invitrogen) was used as the cloning and the expression host strain using the pET20b vector (Novagen). The procedure for obtaining the gene for producing the type I AFP genes was previously described by Solomon and Appels.21 We used a slightly modified version of their technique to obtain the sequence for rAFP (Table 1). Briefly, four primers were used, two of which included the major portion of the sequence that was extended using the two other shorter primers according to the overlap extension technique. It is noteworthy that rAFP was designed based on 11 amino acid repeat sequences found in the native winter flounder proteins, but differs in that the repeat sequences have less variability than the native proteins and it contains five 11 residue repeats instead of the three or four repeats naturally found. In order to conjugate this protein to a polyallylamine (PAA) specifically through its carboxy terminus, we produced a variant of the protein with the two aspartic acid residues replaced by asparagines, which have a similar side group size but without the carboxylic acid group (Table 1). This modified protein, rAFP-N, was prepared with the same procedure to create the gene for producing rAFP except using a slightly modified forward extension primer to replace the aspartic acid residues with asparagines. PCR amplifications were carried out using a thermal cycler (Thermo Electron Corp.) in a 50 μL reaction containing 0.01 μM template primers, 1.25 μM short primers, nuclease free water, and PCR master mix (Promega). Purified DNA and the pET20b expression vector were each 10-fold over digested with NdeI and BamHI restriction endonucleases. Digestion products were prepared for ligation by purification with agarose gel electrophoresis.
ARTICLE
Table 1. Amino Acid Sequence and Thermal Hysteresis Activity of Reported AFP Constructs activity at 0.3 mM protein
sequence/structure
(°C)
rAFP
MD TASDAAAAAAL (TAANAAAAAAL)4 TAR
0.28
rAFP-N
MN TASNAAAAAAL (TAANAAAAAAL)4 TAR
rAFP-N/PAA Polyallylamine-graft-rAFP-N
0.23 0.50
Quick Ligase Buffer (New England Biolabs) was used in 5 min ligation reaction including approximately 50 ng of the vector DNA and a 10-fold molar excess of the insert DNA followed by plating. After overnight growth, colonies were selected and plasmid DNA was purified using a miniprep kit (Qiagen). DNA sequencing (Cleveland State University Genomics Laboratory) was used to verify that the gene was ligated into the plasmid properly. Protein Expression and Purification. Overnight cultures of transformed bacteria in 5 mL of ampicillin containing LB medium was used to inoculate 1 L cultures having 100 mg ampicillin. Samples were induced with isopropyl-β-D-thiogalactopyranoside (IPTG) (Thermo Fisher Scientific) after cultures reached an optical density of 0.6 at 600 nm. Induced samples were shaken in an incubator at 37 °C for 6 h. Samples were purified (>95% as measured by polyacrylamide gel electrophoresis and reverse-phase HPLC) using ethanolic extraction according to the protocol reported by Solomon and Appels.21 The purified protein was dialyzed against 100 mM ammonium bicarbonate using 1000 Da cutoff membrane followed by lyophilization. Molecular weights of the proteins were determined by mass spectrometry as 5108.6 and 5105.6 Da for rAFP and rAFP-N, respectively. These results were in accordance with the weights calculated from the primary sequences. Concentration of a master solution was determined by amino acid analysis (University of Oklahoma Molecular Biology Proteomics Facility). Conjugation Reactions. In a round-bottom flask, 5 μL polyallylamine (Sigma-Aldrich) (Mw ∼17 000, 20 wt % in water) was mixed with 24 μL EDC (1-Ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride) (Sigma-Aldrich) and 10.9 μL sulfo-NHS (n-hydroxysulfosuccinimide) (Sigma-Aldrich) in 300 μL 0.4% acetic acid, 30 mM ammonium bicarbonate.22 Approximately 13 times molar excess of the rAFP-N (0.3 mM) was added to the reaction mixture, which was stirred for 2 h at room temperature. Unreacted impurities were removed by dialysis using 3500 MW cutoff Spectra/Por dialysis membrane. Size Exclusion HPLC. A 500 μL aliquot of the reaction mixture was loaded on the 20 mL bioscale size exclusion column (BioRad Laboratories) filled with Toyopearl HW-55F resin (Tosoh). Reaction buffer solution was pumped through the column in the down-flow mode at a flow rate of 1 mL/min. UV absorbance at 230 and 280 nm were used to monitor the eluted sample. Thermal Hysteresis Experiments. The concentration dependence of the thermal hysteresis activity of the AFPs was measured as the difference between the melting and the freezing temperatures of protein solutions. The melting and freezing points were measured as previously described using a homebuilt nanoliter osmometer attached to an optical microscope (Olympus).23 Briefly, a single drop of protein solution was suspended in oil. 2167
dx.doi.org/10.1021/bc2004318 |Bioconjugate Chem. 2011, 22, 2166–2171
Bioconjugate Chemistry
ARTICLE
The droplet was flash frozen at around 30 °C and, upon increasing the temperature to the melting point, was thawed slowly until a single ice crystal was obtained. The temperature was subsequently decreased (0.01 °C/min) until reaching the freezing temperature, where rapid crystal growth was observed.
’ RESULTS We successfully expressed and purified two different recombinant type I antifreeze proteins, rAFP and rAFP-N (Table 1), as verified by DNA sequencing of the gene, mass spectrometry, and amino acid analysis. The rAFP protein was first designed and expressed by Solomon and Appels,21 based on the HPLC6 winter flounder type I AFP. It consists of initial HPLC6 residues, including the first 11-residue repeat sequence, followed by four of the terminal HPLC6 11-residue repeats. The five 11-residue repeats is greater than the three commonly found in nature. The rAFP-N protein is the same as rAFP except that mutations were introduced to substitute the two aspartic acid residues with asparagine residues. This eliminates the acidic side groups that could compete with the protein carboxy terminus in the conjugation reaction with the amine groups of the polyallylamine. The conjugation reaction product of rAFP-N (5.1 kD) with PAA (17 kD) was passed through a size exclusion chromatography column (Figure 1) and fractionated to determine which peaks contained active antifreeze protein, as determined by thermal hysteresis measurements and ice crystal morphologies. The 280 nm absorbance is attributed to the PAA, whereas both the protein and PAA absorb at 230 nm. The first (high molecular weight) peak contains active antifreeze protein and corresponds to the reaction product. The second peak also exhibits activity, and is attributed to unreacted protein. None of the other peaks contained active protein. All three protein constructs (rAFP, rAFP-N, and rAFP-N/ PAA) exhibit thermal hysteresis activity (Figure 2). Although rAFP and rAFP-N have comparable activities up to 0.2 mM, the activity for rAFP-N was lower than that for rAFP at higher concentrations. We speculate that the lower thermal hysteresis activity of rAFP-N may be the result of reduced solubility since the charged aspartic acid residues were removed to facilitate the specific conjugation reactions, but it could also be caused by structural differences induced by the point mutations. When compared to reported thermal hysteresis values of its shorter isoform HPLC-6,24 rAFP is more active at all concentrations measured, i.e., up to 2 mM. The activity of the rAFP-N/PAA is measured using the reaction mixture, which includes both the rAFP-N conjugated to the PAA and any unreacted protein. For comparison to the unconjugated protein, the thermal hysteresis is plotted against the protein domain concentration, which includes both protein domains conjugated to the polymer and free unreacted domains. This domain concentration is based on the known amount of protein added in the reaction. At concentrations below 0.3 mM of rAFP-N, the rAFP-N/PAA reaction product is at least twice as active compared to rAFP or rAFP-N. It would have been preferable to compare the conjugation product purified from unreacted protein by chromatography. However, quantification of the protein domain concentration in the fractionated samples was not successful. We attempted using NMR, but the signal from the acetic acid in the sample buffer, which was necessary to solubilize the product, overwhelmed the NMR spectra of the sample. Amino acid analysis of the product also did not produce
Figure 1. Size exclusion chromatograph of the conjugation reaction mixture. The sample was fractionated at several intervals to identify the peaks that contained active antifreeze proteins. The sample is detected by UV absorbance (black line: 230 nm, gray line: 280 nm).
Figure 2. Concentration dependence of thermal hysteresis. Thermal hysteresis measurements of the pure proteins, rAFP (diamonds) and rAFP-N (solid circles), and the reaction product, rAFP-N/PAA (open circles), were measured at a cooling rate of 0.01 °C/min. The concentrations for rAFP-N/PAA indicate the molar concentration of the rAFPN protein molecules used in the reaction mixture.
Figure 3. Ice crystal morphologies in the presence of antifreeze protein constructs: A, rAFP (530 μM); B, rAFP-N (600 μM); C, rAFP-N/PAA (32 μM of rAFP-N). For all samples, the c-axis is parallel to the paper plane. Scale bars are 10 μm.
reliable results presumably because of the interference of the polymer backbone. Ice crystal morphologies were observed in the presence of the AFP. At all concentrations, the ice crystal morphologies in the presence of rAFP were similar, with ice crystals between 10 and 15 μm at low concentrations and slightly larger at 0.53 mM (Figure 3A). Truncated bipyramidal ice crystals are initially 2168
dx.doi.org/10.1021/bc2004318 |Bioconjugate Chem. 2011, 22, 2166–2171
Bioconjugate Chemistry
Figure 4. Ice crystal morphologies in the presence of the HPLC purified rAFP-N/PAA. The c-axis is parallel to the paper plane except for the third row, in which it is perpendicular to the paper plane. From left to right: Initial single ice crystal, ice crystal just before and after the burst point, respectively. First row is the left arm of the first peak in Figure 1, while the remaining rows represent the top-right of the first peak, bottom-right of the first peak, and the second peak in Figure 1, respectively. No ice crystal morphology change was detected for the remainder of the peaks. Scale bar is 10 μm.
formed, followed by more complete bipyramid formation over time. Similar hexagonal bipyramidal ice crystal morphologies have been reported in the presence of the shorter isoform of rAFP.25 At the burst point, the ice crystals fail at the bipyramid tip growing in the direction of the c-axis. The rAFP-N also produced bipyramidal structures, which were frequently truncated at lower concentrations. Surprisingly, the rAFP-N burst point morphology was substantially different than rAFP. Ice crystals failed at the edge and center of the crystal with ice crystal growth always along the a-axis (Figure 3B). Ice crystal morphologies of the rAFP-N/ PAA reaction product revealed similar behavior as rAFP-N except that much smaller ice crystals were obtained at all concentrations (Figure 3C). Again, failure occurred at the tip and at the center of the ice crystal growing along the a-axis, similar to rAFP-N. The first and second peaks of the rAFP-N/PAA reaction product purified by size exclusion chromatography (Figure 1) were fractionated and ice crystal morphologies were observed (Figure 4). The concentration for these samples is unknown due to the inability to quantify the conjugated protein adequately. Thermal hysteresis activities of the samples from first row to the fourth row are 0.25, 0.28, 0.20, and 0.11 °C, respectively. No ice crystal morphology change was detected for the remainder of the peaks indicating that the first two peaks constitute the majority of the reacted and unreacted protein.
’ DISCUSSION Activity of antifreeze proteins is primarily measured through the concentration-dependent thermal hysteresis, which is the
ARTICLE
difference between the melting and freezing points of ice crystals in AFP solutions. Thermal hysteresis results from the noncolligative depression of the freezing point through AFP interaction with the ice crystal surface. Several factors have been reported to affect the thermal hysteresis activity of a given protein, including the size of the protein (steric effect), the surface area and lattice matching of the ice binding face (affinity), and the ability of the multiple protein domains to bind with the ice surface simultaneously (avidity).14,15,17,20 The widely accepted mechanism for thermal hysteresis is described by a model based on the Kelvin effect.26 Once proteins bind to the ice surface, the spacing between the adsorbed protein molecules determines the level of thermal hysteresis activity. This model relies on the pinning of the ice crystal by the proteins leading to curvature of the growing ice surface between the protein molecules resulting in an increased energy barrier for the addition of water molecules. This can stop the growth of the ice crystal, effectively lowering the freezing point. Since the growth of the ice crystals has been observed to be completely stopped, the proteins must be irreversibly adsorbed to the ice. Irreversible adsorption conflicts with the observed concentration dependence of thermal hysteresis, which suggests that there is an equilibrium process occurring. To reconcile these observations, Kristiansen and Zachariassen27 presented a convincing model with a two-step process: protein equilibrium (reversible) binding occurs at the colligative freezing temperature as the ice begins to form, but as the crystals mature, the proteins become trapped in the surface (irreversible) to result in thermal hysteresis. The concentration dependence of thermal hysteresis is a result of different protein concentrations present at the interface due to the reversible binding step. On the basis of the above model, for a given antifreeze protein that binds to a specific plane of the ice surface, increasing the concentration at the interface during the reversible adsorption step will tend to increase the thermal hysteresis. Adsorption/ desorption equilibrium should therefore be considered when designing an antifreeze protein molecule. The most common approach used to increase surface concentrations has been to increase the affinity of protein binding to the ice, by increasing the area of the protein binding face for helical AFPs.11,12 The increase in affinity reduces the rate of desorption leading to an increased equilibrium protein concentration at the surface. Increasing the affinity is to some extent limited in that the quality of lattice matching diminishes with the addition of helical repeats. Increased avidity also can result in increased thermal hysteresis, particularly at low protein concentrations, as has been reported with the naturally occurring two-domain type III AFP.20 Using a modified Langmuir adsorption model, we illustrated that this can be explained by an increased surface concentration during reversible adsorption.28 This insight inspired the design of the high-avidity AFP construct produced by the conjugation of multiple AFPs to a polymer chain (Figure 5). The high avidity results in an increased concentration of the surface-bound proteins and activity of the product. Hence, a smaller concentration of protein is required to obtain the same activity compared to free AFP. The proposed mechanism of action is for the increase in activity of the rAFP-N/PAA is the increased avidity. During the reversible step of the AFP interaction with ice crystals, multiple proteins attached to a single polymer chain can adsorb to the ice surface. Even if there was a relatively weak interaction of 2169
dx.doi.org/10.1021/bc2004318 |Bioconjugate Chem. 2011, 22, 2166–2171
Bioconjugate Chemistry
Figure 5. Schematic representations of rAFP-N/PAA reaction product. The rAFP-N models were created by extending the length of winter flounder AFP (PDB ID 1WFA37) while maintaining the α-helical structure, using DeepView/Swiss-PdbViewer software program.38 PAA (17 kDa) was produced using PyMOL version 0.99.
the individual proteins, resulting in significant probability of exchange of the molecules, desorption of one or more of the protein molecules would not be enough to disengage the whole construct from the surface. All proteins would need to desorb simultaneously, leading to near-zero probability of complete desorption. This will lead to high surface coverage even at low solution concentrations. It is noteworthy that, using a similar approach, the conjugation of hydrophobic side chains to polymers has been used to produce stable adsorbed layers on hydrophobic surfaces.29 As expected, in the reaction product containing the AFP conjugated to the PAA we observed an increase in thermal hysteresis compared to equivalent concentrations of pure protein, particularly at low concentrations. We expect that using reaction mixture containing both the conjugate and unreacted protein underestimates the actual thermal hysteresis values of conjugation product with no free AFP. The degree of this underestimation may be relatively small in light of the reported conferring of full activity to a defective AFP by the addition of a small fraction of an active AFP.30 Even with the potential underestimation of the effectiveness, it is impressive that the conjugated product was able to achieve a thermal hysteresis of 0.3 °C with about 1% of the total protein concentration needed for the free AFP (3 μM of the protein domain compared to 300 μM, respectively). This drastically lowers the total amount of protein necessary to achieve the same activity. If the high avidity of the rAFP-N/PAA leads to high surface coverage particularly at low concentrations, one may question why a greater increase in thermal hysteresis is not observed. Although there is significant thermal hysteresis activity improvement within the concentration range studied, these values fall below the thermal hysteresis values reported for the hyperactive type I AFP.13 A likely answer is that the distribution of the protein along the polymer both in distance and in frequency may not be optimal for maximizing surface coverage. This could lead to incorrect alignment of the protein molecules on the ice lattice. The reaction product obtained in this study has multiple protein molecules attached to a polyamine carrying multiple ice binding sites. It is likely that misalignment of the ice binding faces during the reversible exchange is more pronounced for this bigger molecule compared to individual protein molecules and hence there might be a limit to the number of protein molecules on the polymer chain that can simultaneously align with the ice surface.
ARTICLE
By changing the ratio of protein to polymer, an optimal number of proteins for the maximizing thermal hysteresis activity could be obtained. If greater thermal hysteresis is desired, another approach could be to use more active antifreeze proteins such as insect antifreeze proteins. Different types of AFPs produce different ice crystal morphologies, which is attributed to their structural differences.31 For example, type I AFPs bind to the bipyramidal (201) plane of ice.26 Molecular modeling studies revealed that the main source of this interaction was the steric compatibility between the ice plane and the AFP.32 Type III AFPs comparably bind to the prism (100) plane of the ice crystal.33 Similarly, steric contributions and hydrophobic interactions were found to play an important role in ice binding of these AFPs by using molecular dynamics simulations, ice docking, and energy minimization studies.26,34 36 The ice crystal morphology observed for both rAFP-N and the conjugation product rAFP-N/PAA is initially a truncated bipyramid that may grow with time to form a more complete bipyramid, more often at higher concentrations. The similar behavior suggests that the conjugation does not alter the manner in which the proteins bind to the ice surface. It is unknown why the morphology of the ice crystal growth is different between the rAFP and the rAFP-N. The conversion of the two aspartic acids to asparagines are outside of the binding region of the protein and are not expected to alter the simple α-helical fold of the protein. For some unknown reason, the loss of the charged residues leads to the greater prevalence of truncated bipyramidal crystals, as well as burst growth along the a-axis instead of the c-axis. Creating superior AFP constructs that have thermal hysteresis activity at low concentrations would make it more feasible economically to use them as additives for cryopreservation and refrigeration systems. We have successfully demonstrated a potential route toward this goal. Through the use of higheractivity AFP domains, such as insect AFPs, and optimizing the density of conjugation, even greater improvements in activity might be obtained.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses †
Department of Medical Biochemistry, School of Medicine, Acibadem University, Istanbul, Turkey.
’ ACKNOWLEDGMENT We thank A. H. Heuer and A. McIlwain (Case Western Reserve University) for use of and assistance with the nanoliter osmometer and M. A. Ruegsegger (The Ohio State University) for helpful advice with the bioconjugation procedure. This research is supported by awards from the American Heart Association (Scientist Development Grant, 0635084N) and the Cleveland State University Doctoral Dissertation Research Expense Award Program. ’ ABBREVIATIONS AFP, antifreeze protein; PAA, polyallylamine; rAFP, recombinant type I fish AFP; rAFP-N, rAFP with aspartic acid residues replaced by asparagine; rAFP-N/PAA, polyallylamine-graftrAFP-N 2170
dx.doi.org/10.1021/bc2004318 |Bioconjugate Chem. 2011, 22, 2166–2171
Bioconjugate Chemistry
’ REFERENCES (1) Venketesh, S., and Dayananda, C. (2008) Properties, potentials, and prospects of antifreeze proteins. Crit. Rev. Biotechnol. 28, 57–82. (2) Knight, C. A., Wen, D. Y., and Laursen, R. A. (1995) Nonequilibrium antifreeze peptides and the recrystallization of ice. Cryobiology 32, 23–34. (3) Amir, G., Horowitz, L., Rubinsky, B., Yousif, B. S., Lavee, J., and Smolinsky, A. K. (2004) Subzero nonfreezing cryopreservation of rat hearts using antifreeze protein I and antifreeze protein III. Cryobiology 48, 273–282. (4) Soltys, K. A., Batta, A. K., and Koneru, B. (2001) Successful nonfreezing, subzero preservation of rat liver with 2,3-butanediol and type I antifreeze protein. J. Surg. Res. 96, 30–34. (5) Chao, H. M., Davies, P. L., and Carpenter, J. F. (1996) Effects of antifreeze proteins on red blood cell survival during cryopreservation. J. Exp. Biol. 199, 2071–2076. (6) Inada, T., and Modak, P. R. (2006) Growth control of ice crystals by poly(vinyl alcohol) and antifreeze protein in ice slurries. Chem. Eng. Sci. 61, 3149–3158. (7) Egolf, P. W., and Kauffeld, M. (2005) From physical properties of ice slurries to industrial ice slurry applications. Int. J. Refrig. 28, 4–12. (8) Griffith, M., and Ewart, K. V. (1995) Antifreeze proteins and their potential use in frozen foods. Biotechnol. Adv. 13, 375–402. (9) Yang, D. S. C., Sax, M., Chakrabartty, A., and Hew, C. L. (1988) Crystal-structure of an antifreeze polypeptide and its mechanistic implications. Nature 333, 232–237. (10) Patel, S. N., and Graether, S. P. (2010) Structures and icebinding faces of the alanine-rich type I antifreeze proteins. Biochem. Cell Biol. 88, 223–229. (11) Chao, H., Hodges, R. S., Kay, C. M., Gauthier, S. Y., and Davies, P. L. (1996) A natural variant of type I antifreeze protein with four icebinding repeats is a particularly potent antifreeze. Protein Sci. 5, 1150–1156. (12) Gauthier, S. Y., Marshall, C. B., Fletcher, G. L., and Davies, P. L. (2005) Hyperactive antifreeze protein in flounder species. FEBS J. 272, 4439–4449. (13) Marshall, C. B., Chakrabartty, A., and Davies, P. L. (2005) Hyperactive antifreeze protein from winter flounder is a very long rodlike dimer of alpha-helices. J. Biol. Chem. 280, 17920–17929. (14) Leinala, E. K., Davies, P. L., Doucet, D., Tyshenko, M. G., Walker, V. K., and Jia, Z. C. (2002) A beta-helical antifreeze protein isoform with increased activity - Structural and functional insights. J. Biol. Chem. 277, 33349–33352. (15) Marshall, C. B., Daley, M. E., Sykes, B. D., and Davies, P. L. (2004) Enhancing the activity of a beta-helical antifreeze protein by the engineered addition of coils. Biochemistry 43, 11637–11646. (16) Miura, K., Ohgiya, S., Hoshino, T., Nemoto, N., Suetake, T., Miura, A., Spyracopoulos, L., Kondo, H., and Tsuda, S. (2001) NMR analysis of type III antifreeze protein intramolecular dimer - Structural basis for enhanced activity. J. Biol. Chem. 276, 1304–1310. (17) Baardsnes, J., Kuiper, M. J., and Davies, P. L. (2003) Antifreeze protein dimer - When two ice-binding faces are better than one. J. Biol. Chem. 278, 38942–38947. (18) Nishimiya, Y., Ohgiya, S., and Tsuda, S. (2003) Artificial multimers of the type III antifreeze protein - Effects on thermal hysteresis and ice crystal morphology. J. Biol. Chem. 278, 32307–32312. (19) Holland, N. B., Nishimiya, Y., Tsuda, S., and S€onnichsen, F. D. (2007) Activity of a two-domain antifreeze protein is not dependent on linker sequence. Biophys. J. 92, 541–546. (20) Holland, N. B., Nishimiya, Y., Tsuda, S., and S€onnichsen, F. D. (2008) Two domains of RD3 antifreeze protein diffuse independently. Biochemistry 47, 5935–5941. (21) Solomon, R. G., and Appels, R. (1999) Stable, high-level expression of a type antifreeze protein in Escherichia coli. Protein Express. Purif. 16, 53–62. (22) Hermanson, G. T. (2008) Bioconjugate Techniques, 2nd ed., Academic Press, Burlington, MA. (23) Miller, N. E. (2003) “Burst-Point” Morphologies of Ice Crystals in the Presence of Antifreeze Proteins, Masters Thesis, Department of
ARTICLE
Materials Science and Engineering, Case Western Reserve University, Cleveland, OH. (24) Baardsnes, J., Kondejewski, L. H., Hodges, R. S., Chao, H., Kay, C., and Davies, P. L. (1999) New ice-binding face for type I antifreeze protein. FEBS Lett. 463, 87–91. (25) Evans, R. P., and Fletcher, G. L. (2001) Isolation and characterization of type I antifreeze proteins from Atlantic snailfish (Liparis atlanticus) and dusky snailfish (Liparis gibbus). Biochim. Biophys. Acta 1547, 235–244. (26) Knight, C. A., Cheng, C. C., and Devries, A. L. (1991) Adsorption of alpha-helical antifreeze peptides on specific ice crystalsurface planes. Biophys. J. 59, 409–418. (27) Kristiansen, E., and Zachariassen, K. E. (2005) The mechanism by which fish antifreeze proteins cause thermal hysteresis. Cryobiology 51, 262–280. (28) Can, O., and Holland, N. B. (2009) Modified Langmuir isotherm for a two-domain adsorbate: Derivation and application to antifreeze proteins. J. Colloid Interface Sci. 329, 24–30. (29) Holland, N. B., Qiu, Y. X., Ruegsegger, M., and Marchant, R. E. (1998) Biomimetic engineering of non-adhesive glycocalyx-like surfaces using oligosaccharide surfactant polymers. Nature 392, 799–801. (30) Takamichi, M., Nishimiya, Y., Miura, A., and Tsuda, S. (2009) Fully active QAE isoform confers thermal hysteresis activity on a defective SP isoform of type III antifreeze protein. FEBS J. 276, 1471–1479. (31) Chao, H., Deluca, C. I., and Davies, P. L. (1995) Mixing antifreeze protein types changes ice crystal morphology without affecting antifreeze activity. FEBS Lett. 357, 183–186. (32) Lal, M., Clark, A. H., Lips, A., Ruddock, J. N., and White, D. N. J. (1993) Inhibition of ice crystal-growth by preferential peptide adsorption: A molecular modeling study. Faraday Discuss. 95, 299–306. (33) Madura, J. D., Baran, K., and Wierzbicki, A. (2000) Molecular recognition and binding of thermal hysteresis proteins to ice. J. Molec. Recog. 13, 101–113. (34) Haymet, A. D. J., Ward, L. G., and Harding, M. M. (1999) Winter flounder “antifreeze” proteins: Synthesis and ice growth inhibition of analogues that probe the relative importance of hydrophobic and hydrogen-bonding interactions. J. Am. Chem. Soc. 121, 941–948. (35) Loewen, M. C., Chao, H. M., Houston, M. E., Baardsnes, J., Hodges, R. S., Kay, C. M., Sykes, B. D., S€onnichsen, F. D., and Davies, P. L. (1999) Alternative roles for putative ice-binding residues in type I antifreeze protein. Biochemistry 38, 4743–4749. (36) S€onnichsen, F. D., DeLuca, C. I., Davies, P. L., and Sykes, B. D. (1996) Refined solution structure of type III antifreeze protein: Hydrophobic groups may be involved in the energetics of the protein-ice interaction. Structure 4, 1325–1337. (37) Sicheri, F., and Yang, D. S. C. (1995) Ice-binding structure and mechanism of an antifreeze protein from winter flounder. Nature 375, 427–431. (38) Guex, N., and Peitsch, M. C. (1997) SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophor. 18, 2714–2723.
2171
dx.doi.org/10.1021/bc2004318 |Bioconjugate Chem. 2011, 22, 2166–2171
