Discovery and characterization of an antifreeze protein from Lake Ontario midges.

摘要

This thesis describes the discovery and characterization of the first antifreeze protein (AFP) isolated from a fly. The starting point of my research was the observation that when midges emerge from Lake Ontario as adults in early spring they have low levels of antifreeze activity. Here I have isolated, characterized, and modeled the structure of their main AFP. This 79-residue mature midge AFP has a novel sequence of ten-residue tandem repeats, xxCxGxYCxG, which I modeled as is a left-handed disulfide-braced solenoid, with each 10-residue repeat corresponding to one coil of the helix. The fold is similar to a beta-helix, but secondary structure and circular dichroism analyses indicate that this solenoid is too tightly coiled to have beta-structure. The model shows an outward pointing seven-residue stacked Tyr-ladder, which has been confirmed by mutagenesis to serve as its ice-binding site. This is the first example of tyrosines used for ice binding. I have determined that the midge AFP activity is intermediate to the moderately active and hyperactive AFPs typically found in fish and overwintering insects, respectively. The proposed explanation for intermediate activity is that the midge AFP binds to a pyramidal surface midway between the basal and prism planes. The modest sub-zero temperatures that the adult flies encounter has likely driven the evolution of this intermediate activity AFP. I predict that other organisms facing freezing threats of a similar magnitude will also produce AFPs with intermediate activity. I have also contributed to the literature on experimental methods used to assess ice-binding properties of AFPs by publishing a step-by-step protocol for the fluorescence-based ice-plane affinity assay with an instructional video. This technique can be used, as it was here, to determine which planes of ice are bound by fluorescently-labelled AFPs. A second chapter on techniques describes the ab initio and homology-based techniques our lab has used to reliably predict novel ice-binding protein folds.