| dc.description.abstract | Antifreeze proteins (AFPs) are produced by a variety of so-called freeze-avoiding animals, ectotherms that adaptively allow their body fluids to supercool as a response to subfreezing temperatures in their environment. AFPs are defined by their ability to prevent the growth of ice. The separation between the melting -and freezing temperature of ice is referred to as thermal hysteresis, and its strength is called the antifreeze activity. Thermal hysteresis results from the AFPs adsorbing to the ice surface, and the antifreeze activity is dependent on the ice surface density of adsorbed AFPs. The antifreeze activity may be elevated when the AFP interact with other proteins, so-called enhancers. The enhancers do not themselves have the ability to cause thermal hysteresis. Certain observed characteristics of thermal hysteresis seem to be in conflict with predictions from the current theoretical explanation to the phenomenon. This investigation consists of five individual studies (I – V) with two overall aims: First, to expand the theoretical description of how AFPs cause thermal hysteresis to explain the apparent inconsistencies between theory and observations. Second, to study the causes and effects of AFPs interacting with other proteins, expressed as enhancement of the antifreeze activity.
The AFP used in this investigation was isolated from the hemolymph of the longhorn beetle Rhagium inquisitor (I). This AFP, denoted RiAFPH4, is only one of at least six AFP forms within the hemolymph of this species, and is found in much higher levels than the other forms (I). Its primary sequence is quite different from any other insect AFPs in several respects (II): It contains seven repeats of a unique thirteen residue pattern. The repeat regions are not connected in series, as seen in other insect AFPs, but spaced apart by a varying number of residues. The repeat pattern is palindromic, i.e. similar when read in either direction. The structure contains only a single disulfide bond, as opposed to other insect AFPs that contain a number of such bonds. Each repeat region potentially contains three ice - binding motifs. The significance of the palindromic character of the pattern and the absence of central disulfide bonds may be to allow the protein chain to make turns. This would make possible the arrangement of different ice-binding motifs, spaced irregularly apart in the primary sequence, into an internal ice binding site similar to that seen in other insect AFPs.
RiAFPH4 was used in the experimental part of the investigation that was concerned with the antifreeze mechanism (I, II, IV) or the causes and consequences of AFP – protein interactions (IV, V). Study III was an examination and expansion of the theoretical explanation to thermal hysteresis based on available literature data. The antifreeze mechanism AFPs are assumed to cause thermal hysteresis by adsorbing irreversibly to the ice surface, thereby causing the ice to grow as convex growth zones at the exposed areas between the adsorbed AFPs. This is referred to as the adsorption inhibition mechanism. According to this view ice growth inhibition is caused by ice surface water molecules acting as an energy barrier to ice growth. Study III relates thermal hysteresis to ice - water vapour pressure equilibrium. Surface curvatures affect the vapour pressure, referred to as the Kelvin effect, and the convexity of the ice fronts postulated by the adsorption inhibition mechanism elevates the vapour pressure of the ice to that of the surrounding supercooled solution. In this manner an ice – water equilibrium is maintained within a temperature interval below the equilibrium freezing temperature, and the ice crystal consequently does not expand visibly. Thus, study III modifies the theoretical basis for the ice growth inhibition. Irreversible adsorption of the AFPs to the ice surface is a fundamental presupposition of the adsorption inhibition model. If the AFPs desorbs from the ice surface, then the antifreeze effect will disappear. There are a number of apparent inconsistencies between the premise of irreversible adsorption and observations (III). All of these inconsistencies may be explained by assuming that the process that results in the observed antifreeze activity consists of two steps (III). At the melting temperature of the crystal, the AFPs are “melted off” the ice, and an equilibrium distribution of AFPs is established between the ice surface region and the surrounding solution. As the temperature is lowered the AFPs become “frozen” irreversibly to the ice surface. Consistent with this, RiAFPH4 is only reversibly associated to the ice while the antifreeze activity is determined (IV).
The establishment of equilibrium distribution of AFPs while at the melting temperature of the crystal must be affected partly by the solubility of the AFPs in solution and partly by their ability to remain within the surface region. Hence, these two properties of the AFP are assumed to determine the antifreeze activity. Characteristics known to affect the solubility of proteins in solution also affect the antifreeze activity (III). The ability of RiAFPH4 to induce antifreeze activity depends on the pH of the solution (I, V). The hydrophobic character of proteins depends on their charge, and protein charge is pH - dependent. The observed effect of pH on the antifreeze activity is consistent with a pH – induced change in the solubility of the AFP that affects the equilibrium distribution of RiAFPH4 between ice and water. RiAFPH4 appear to be the most potent AFP discovered to date (II). Its primary structure suggests that it contains about twice the number of ice binding motifs as seen in other insect AFPs (II). These ice - binding motifs may cause RiAFPH4 to remain within the interfacial region prior to irreversible adsorption by forming transient hydrogen bonds with the more solid-like deeper part of the ice surface region. This would shift the equilibrium distribution of this AFP further towards the ice and hence, result in increased antifreeze activity. AFP - protein interactions
Current theory as to how enhancers cause an elevation of the antifreeze activity is that they reduce the spacing between the adsorbed AFPs on the ice surface by increasing the size of the adsorbed structure. It was shown in study IV that the observed enhancement of RiAFPH4 by albumin is consistent with the enhancer reducing the solubility of the AFP in solution, thereby shifting the equilibrium distribution of AFPs more towards the ice surface. This would in turn result in greater surface densities of adsorbed AFPs and hence increase the antifreeze activity, in accordance with the concept of equilibrium distribution of AFPs between the ice surface and solution proposed in III. Hence, an alternative explanation to the phenomenon of enhancement is provided.
The hemolymph AFPs of R. inquisitor appear to be enhanced in vivo (V). Chemical interaction between an AFP and other proteins will in principle take place only if they are not repelled from each other by having similar electrical charges. This implies that interactions can occur only when the pH value of the solution is between the isoelectric points of the AFPs and the enhancer. RiAFPH4 (isoelectric point ~ 8.5) was enhanced by both albumin (isoelectric point 4.8) and carbonic anhydrase (isoelectric point 5.6), but only within a pH - interval corresponding quite closely to that between their respective isoelectric points, suggesting that the net charge of AFPs and enhancers is an underlining determinant for the AFP – enhancer interaction.
All hemolymph AFPs of R. inquisitor seem to be cationic at physiological pH (6.7), with isoelectric points > 8 (I). This is also the case for the bulk of the hemolymph proteins (V). Hence, the hemolymph AFPs of R. inquisitor seem to be inhibited from interacting with the majority of hemolymph proteins by its net electric charge in vivo.
In addition to the requirement of opposite net charge, the interaction between AFPs from R. inquisitor and enhancers seem to depend on specific surface charges (V). Hence, the AFP – protein interaction could be avoided by the AFP not having the required charged residues for the interaction to occur. The observed enhancement of hemolymph AFPs of R. inquisitor (V) is therefore a strong indication that the AFP – enhancer interaction is functional.
The apparent dependency of the AFP – protein interaction on specific surface charges may provide an explanation to the high number of AFP isoforms seen in insects (e.g. I), their differentiated expression (e.g. I) and specific distribution within body fluid compartments; different AFP forms may have different arrangement of surface charges to optimize interaction with specific proteins. Thus, the primary function of the AFP – enhancer interaction may not be the observed enhancement effect, but be directed towards the protein the AFP interacts with, such as stabilization of protein structure during exposure to low temperature (V). This would be analogous to the role played by heat shock proteins. The AFP – ice interaction is independent of the net charge of the AFP (I, V). Hence, the charge features of AFPs seem to be a functional property of their structure, separate from their ability to adsorb to ice. | nb_NO |
