Engineering methionine gamma lyase from the cheese bacterium Brevibacterium aurantiacum for cancer therapy

Master

anglais. The cancer incidence is increasing worldwide, where the disease is the second cause of mortality. The methionine depletion induced by the methionine γ‐lyase enzyme (MGL, methioninase) is a promising strategy against cancer, whose cells are more dependent on methionine than normal cells. MGL is a pyridoxal phosphate (PLP) dependent enzyme that degrades methionine. It is present in bacteria, in plants, but absent in mammals. This work studied the methioninase from Brevibacterium aurantiacum (Ba-MGL). This enzyme from a cheese bacterium should comply with human immune tolerance. In addition, it has the advantage to be specific of methionine, sparing cysteine whose depletion could induce undesirable side effects. However, it is quickly inactivated in human plasma, due to PLP leakage. The aim was to improve by enzymatic engineering the affinity of Ba-MGL for PLP, in order to extend its activity in human blood so that it could be used in anticancer therapy. Quadruple (M97I+M218T+T92S+F187Y) and double (M97I+M218T) mutants, designed to mimic the MGL from Pseudomonas putida, were produced in sufficient quantities and purified. According to UV-visible spectra, these mutants did not retain enough PLP, as was initially expected. Thus, they were not assayed for dialysis. A second series of variants (Ba-MGL M94L and M94I), based on sequence comparison with various bacterial MGLs and human cystathionine γ‐lyase, were produced and purified for dialysis assays. In parallel, we studied the MGL from the model plant Arabidopsis thaliana (At-MGL), which, as Ba-MGL, does not degrade cysteine. Crystallization tests were performed on the quadruple mutant and At-MGL. We obtained first needle-shaped crystals. Overall, five variants were successfully expressed and purified. Our results subsequently obtained show the limits of designing enzyme variants merely from sequence and 3D structure comparisons. In cooperation with a bioinformatics laboratory, new designed mutations are underway to improve stability of Ba-MGL. Crystallization trials of variants from Ba-MGL and At-MGL are also underway, in order to solve the structure of MGL, which would help new engineering strategies.