Antifreeze Protein Dimer

Abstract

A naturally occurring tandem duplication of the 7-kDa type III antifreeze protein from Antarctic eel pout (Lycodichthys dearborni) is twice as active as the monomer in depressing the freezing point of a solution. We have investigated the basis for this enhanced activity by producing recombinant analogues of the linked dimer that assess the effects of protein size and the number and area of the ice-binding site(s). The recombinant dimer connected by a peptide linker had twice the activity of the monomer. When one of the two ice-binding sites was inactivated by site-directed mutagenesis, the linked dimer was only 1.2 times more effective than the monomer. When the two monomers were linked through a C-terminal disulfide bond in such a way that their two ice-binding sites were opposite each other and unable to engage the same ice surface simultaneously, the dimer was again only 1.2 times as active as the monomer. We conclude from these analyses that the enhanced activity of the dimer stems from the two ice-binding sites being able to engage to ice at the same time, effectively doubling the area of the ice-binding site.Fish are protected from freezing by antifreeze proteins (AFPs), 1 which bind to the surface of nucleating ice crystals in their body fluids, thereby reducing their freezing point below that of the ocean (1-3). AFPs create a local curvature of the ice between adsorbed AFPs, which makes it energetically unfavorable for liquid water to join the ice surface (4). This process, known as the Kelvin effect, produces a non-equilibrium reduction of the freezing point of ice below the melting point (5). The difference between the melting and freezing points is the thermal hysteresis gap, and it is within this temperature range that ice growth is prevented and fish are protected from freezing.Type III AFPs belong to one of several structurally distinct antifreeze protein families found in fishes (6 -8). These 7-kDa proteins have a compact ␤-stranded structure (9 -13) that shows homology to the C-terminal domain of sialic acid synthase (14). Structure-function studies have localized the icebinding residues to a flat, amphipathic surface on the protein (11,15,16). This ice-binding face includes several conserved hydrophilic residues (Gln 9 , Asn 14 , Thr 15 , Thr 18 , Gln 44 ) that potentially form hydrogen bonds with water molecules on the ice surface (10, 17), flanked by hydrophobic residues (Leu 10 , Ile 13 , Leu 19 , Val 20 , Val 41 ) on the periphery. Altogether, these residues are thought to make favorable van der Waals contacts with the ice surface (10, 12, 18 -20). In addition, there may also be stabilizing entropic effects from bringing this somewhat hydrophobic surface into contact with ice (10, 21). Type III AFP was first reported to bind to the primary prism plane {10-10} of ice (22 Approximately 20 type III isoforms from five species of zoarcid fishes have now been sequenced, with an overall 50% sequence identity. One isoform (designated RD3) from the Antarctic eel pout (Lycodichthys dearborni) has two ...

A two-dimensional adsorption kinetic model for thermal hysteresis activity in antifreeze proteins

“…53 On the other hand, it is interesting that the ratio of the maximum activities for ␤-helical TmAFP and CfAFP is approximately equal to the ratio of the ice crystal area covered by them. This result is in good coincidence with the viewpoint of Baardsnes et al 52,53 In addition, although there is much difference in antifreeze activity for ␤-helical AFPs, the difference in the number of their ice-binding sites is small as compared to the large difference in TH activity. This result is consistent with the recent experimental results on partitioning of fish and insect antifreeze proteins into ice.…”

“…This result is in agreement with the experimental conclusion obtained by producing recombinant analogs of the linked dimer that assess the effects of protein size and the number and area of the ice-binding site͑s͒. 52 There is increasing evidence showing that antifreeze activity can be enhanced by increasing the length/area of the icebinding site. The type I AFP9 isoform from winter flounder serum at 52 amino acids in length ͑4.3 kDa͒ has an extra 11-amino acid repeat compared with a common serum isoform, HPLC-6, and therefore has a larger ice-binding face.…”

Hyperactive Antifreeze Protein from Winter Flounder Is a Very Long Rod-like Dimer of α-Helices

“…In trying to account for the superior TH activity of this protein, it is clear that specific activity increases with the size of the AFP (26), particularly when it also increases the surface area of the ice binding site (11,20,42,43). This is the general trend in both insect and fish AFPs.…”

“…In fish, the larger antifreeze glycoproteins (AFGPs 1-5) are considerably more active than the shorter ones (AFGPs 6 -8) (44). Tandem repeats (referred to as dimers) of type III AFP, both natural and artificial, show significantly increased TH activity (43,45), particularly when the additional AFP domain can simultaneously engage the ice surface. Even within type I AFPs, the AFP9 isoform that has an extra (fourth) 11-amino acid repeat has noticeably higher activity than the more abundant three-repeat isoforms (11).…”

Modified Langmuir isotherm for a two-domain adsorbate: Derivation and application to antifreeze proteins

“…In the present study, Langmuir adsorption isotherm models are used to relate thermal hysteresis and surface coverage values of single-domain and two-domain type III AFP using previously published thermal hysteresis versus concentration data [38]. Surface coverage and binding energy are determined by fitting two independent thermal hysteresis data sets with the classical Langmuir isotherm and a modified isotherm for the one-and two-domain proteins, respectively.…”