Molecular analysis of the dimerization and aggregation processes of human alanine:glyoxylate aminotransferase and effect of mutations leading to Primary Hyperoxaluria Type I

Primary Hyperoxaluria Type 1 (PH1) is a rare autosomal recessive disorder characterized by the deposition of insoluble calcium oxalate crystals at first in the kidneys and urinary tract and then in the whole body. PH1 is caused by the deficiency of human liver peroxisomal alanine:glyoxylate aminotransferase (AGT). AGT is a pyridoxal 5'-phosphate (PLP)-dependent enzyme, which converts glyoxylate to glycine, thus preventing glyoxylate oxidation to oxalate and calcium oxalate formation. Only two curative therapeutic approaches are currently available for PH1: the administration of pyridoxine (PN), a precursor of PLP, which is only effective in a minority of patients (25- 35%), and liver transplantation, a very invasive procedure. AGT is encoded by the AGXT gene, which is present in humans as two polymorphic forms: the major allele (encoding AGT-Ma) and the minor allele (encoding AGT-Mi). PH1 is a very heterogeneous disease with respect to the clinical manifestations, the response to treatment and the pathogenic mechanisms. In fact, more than 200 pathogenic mutations have been identified so far and the molecular mechanisms by which missense mutations cause AGT deficiency span from functional, to structural and to subcellular localization defects or to a combination of them. Several lines of evidence at both molecular and cellular level, indicate that many disease-causing missense mutations interfere with AGT dimer stability and/or aggregation propensity. However, neither the dimerization nor the aggregation process of AGT have been analyzed in detail. Therefore, we engineered a mutant form of AGT stable in solution in the monomeric form and studied its biochemical properties and dimerization kinetics. We found that monomeric AGT is able to bind PLP and that the coenzyme stabilizes the dimeric structure. Moreover, the identification of key dimerization hot-spots at the monomer-monomer interface allowed us to unravel the mechanisms at the basis of the aberrant mitochondrial mistargeting of two of the most common PH1-causing variants. We also elucidated the molecular and cellular consequences of the pathogenic mutations R36H, G42E, I56N, G63R and G216R, involving residues located at the dimer interface, and tested their in-vitro responsiveness to the treatment with PN. The latter results allowed us to suggest a possible correlation between the structural defect of a variant and its degree of responsiveness to PN. Finally, by combining bioinformatic and biochemical approaches, we analyzed in detail the tendency of AGT to undergo an electrostatically-driven aggregation. We found that the polymorphic changes typical of the minor allele have opposite effect on the aggregation propensity of the protein, and we predicted the possible effect/s of pathogenic mutations of residues located on the AGT surface. Overall, the results obtained allow not only to better understand PH1 pathogenesis, but also to predict the response of the patients to the available therapies as well as to pave the way for the development of new therapeutic strategies.