3-Hydroxy-3-methylglutaryl-coenzyme A lyase: investigation of cysteines mediating intersubunit disulfide formation and regulation by thiol/disulfide exchange and discovery of an extramitochondrial homolog

Abstract

3-Hydroxy-3-methylglutaryl-Coenzyme A lyase catalyzes the cleavage of 3-hydroxy- 3-methylglutaryl-Coenzyme A into acetoacetate and acetyl-Coenzyme A, a key reaction in ketogenesis and leucine catabolism. Previous work prompted the hypothesis that in the absence of reductant, cysteine 323 forms a disulfide bond with cysteine 323 on the adjacent monomer, blocking the substrate's access to the active site which results in diminished enzyme activity. The recently published crystal structure of the human enzyme indicates that cysteines 323 on each monomer are too far apart to form a disulfide. Each of the eight cysteines has been individually mutated to serine and the mutants analyzed for activity and dimer formation in the presence and absence of reductant. C170S, C266S, and C323S do not form dimers in the absence of reductant. C170S, C197S, and C234S have an inflated dependence on thiol, C174S, C307S, and Therefore intersubunit disulfide bond formation does not directly correlate with diminution of activity in the absence of thiol as previously proposed. Analysis of C170S/C174S, C170S/C266S and C170S/C323S double mutants demonstrate that C170S is not directly involved in the intersubunit disulfide bond and suggest that C170 may be involved in regulation of activity by thiol/disulfide exchange. A C266S/C323S mutant heterodimer restores covalent dimer formation confirming that these residues are involved in intersubunit disulfide bond formation. Based on sequence homology, the Mammalian Gene Collection Program identified the protein encoded by the gene 3- hydroxy-3-methylglutaryl-Coenzyme A lyase-like 1 as being a potential 3-hydroxy-3- methylglutaryl-Coenzyme A lyase. Characterization of the purified recombinant human protein confirms that it is an authentic 3-hydroxy-3-methylglutaryl-Coenzyme A lyase. In-vitro myristoylation experiments confirm that the protein is modified by Nmyristoyltransferase and immunofluorescence microscopy in COSI cells demonstrate that this modification effects the subcellular localization of the protein. Blots of rodent organ lysates reveal expression primarily in the small intestine and that the protein is overexpressed in human neuroblastoma and glioma cell lines compared to mouse brain. Co-localization experiments failed to determine to which compartment the lyase-like protein is localized but confirm it is not mitochondrial or peroxisomal as is the traditional lyase.