Protein folding variants, for better or for worse? The lesson learned from HAMLET

Sammanfattning: By varying its secondary or tertiary structure a single protein can attain new biological functions. The amyloid forming proteins are the best known examples where a change in conformation leads to the formation of amyloid fibrils that accumulate in tissues and cause disease. a-lactalbumin in contrast, is an example of a protein that acquires a beneficial function when the conformation is altered. The normal function of the protein is to ac as a substrate specifier for the enzyme galactosyl transferase aiding in the production of lactose. If partially unfolded and stabilized by a specific fatty acid cofactor a-lactalbumin forms HAMLET (Human a-lactalbumin made lethal to tumour cells) that induces apoptosis in tumour cells while sparing healthy cells. We isolated this novel form of a-lactalbumin from human casein and demonstrated that the protein is stable in a partially unfolded conformation even at physiological conditions (paper I). Native a-lactalbumin can be converted to the apoptosis inducing protein if partially unfolded and in the presence of a specific fatty acid cofactor (paper II). The interaction between the protein and the lipid is stereo-specific and we have demonstrated that the fatty acid cofactor have to be unsaturated and in the cis conformation (paper III). The partially unfolded conformation per se is unstable at the conditions used in the apoptosis assay, to investigate whether the protein alone is sufficient to induce apoptosis we used a-lactalbumin mutated in the calcium-binding site thus locked in a partially unfolded conformation. We demonstrate that partially unfolded a-lactalbumin is unable to induce apoptosis thus the conformation and the biological activity of HAMLET is defined both by the protein and by the lipid (paper IV). Finally based on differences in the a-lactalbumin sequence and structure between different species and their ability to form HAMLET like complexes we tentatively identify the fatty acid binding site as being the cluster of hydrophobic residues located in the interface between the a- and the b-domain (paper V). We have thus identified two key elements that regulate the conformation and the function of a-lactalbumin i.e. the calcium ion and the fatty acid cofactor. We speculate that folding variants with new biological functions compared to the native protein may occur very frequently and these mechanisms exist for the benefit of the organism.