Mechanistic studies of the pyridoxal 5'- phosphate-dependent enzyme serine palmitoyltransferase; substrates, cofactor and inhibitors.

Abstract

Sphingolipids (SL) are essential structural components of membranes found in all eukaryotes and have also been identified in some bacteria. The first step of the SL biosynthetic pathway across all species is catalysed by serine palmitoyltransferase (SPT), a member of the alpha-oxoamine synthase (AOS) family of pyridoxal 5’- phosphate (PLP)-dependent enzymes. AOS enzymes are involved in the biosynthesis of a range of important natural products such as heme, vitamins and antibiotics where they catalyse the reaction between amino acid and acyl-thioester substrates. Substrate specificity across the family is of great importance, as human mutant SPTs shift the substrate specificity from L-serine to glycine or L-alanine that lead to production of deoxy-sphingolipids that are toxic to mammalian cells. PLP, a form of vitamin B6, is one of nature’s most versatile catalysts and is involved in over 160 enzymes that carry out diverse reactions on amine-containing substrates. This work probes the functional role of the phosphate group of PLP, usually housed in a phosphate binding cup (PBC) and investigates the need for a novel and unexpected H-bond between the hydroxyl group of the L-serine substrate and the 5’-phosphate group of PLP in SPT. In this study, the PLP cofactor was removed from SPT with amino-thiol substrates which act as mechanism-based inhibitors of SPT via production of a thiazolidine adduct. Replacement of natural PLP with the dephosphorylated form of the cofactor, pyridoxal, allowed a study on the importance of the PLP phosphate:L-serine H-bond on substrate specificity and optimal SPT activity. Furthermore, analysis of the phosphate binding cup of the ALAS:PLP:glycine external aldimine, a related AOS family member; revealed an important residue that could possibly be involved in determining substrate specificity of different members of the AOS family. PBC analysis also expanded, with a detailed and interesting study of a novel SPT:PLP:myriocin inhibitor complex. Human SPT is a heterodimeric, membrane-bound enzyme composed of two subunits (hLCB1/hLCB2) which is thought to contain a single PLP-containing active site. Mutations in human hLCB1 have been linked to the rare sphingolipid metabolic disease hereditary sensory neuropathy I (HSAN1). Recent studies identified three heterozygous missense mutations in the second human SPT subunit hLCB2 that show a significant loss in SPT activity. The three human SPT mutations V359M, G385V and I504F were mapped onto the bacterial S. paucimobilis SPT as V246M, G268V and G385F. These bacterial SPT mutant mimics reveal that the amino acid changes have varying impacts; they perturb the PLP cofactor binding, reduce the affinity for both substrates, decrease the enzyme activity, and, in the most severe case, cause the protein to be expressed in an insoluble form. SPTs and most of the other members of the AOS family utilise an acyl-CoA thioester substrate. In contrast, a sphingolipid-producing bacterium, S. wittichii, is thought to use a small type II acyl carrier protein (ACP) to deliver the acyl chain to its homodimeric SPT target. Converting the unmodified apo-ACP to the activated “substrate” acyl-ACP, has proven difficult and amino acid sequence alignment, combined with modelling studies revealed an unusual tryptophan residue that could prevent modification to the acyl-ACP form. In this study a double mutant ACP E36G/W37A has been prepared and characterised. Both wild-type and mutant S. wittichii ACP are expressed in the recombinant E. coli host in their inactive apoform. The transfer of a phosphopantethiene (4’PP) linker by a specific PPTase (also known as an acyl carrier protein synthase (AcpS)) has been successful in modifying the mutant form of ACP to its holo-form but could not transfer a palmitoyl group (C16). E.coli ACP has been successfully expressed, purified and characterised in this study. For the first time, ion mobility mass spectromerty (IM-MS) has been used on this protein to gain structural insight into the different forms of ACP. Collisional cross section (CCS) distributions have been calculated for different acylated states of the ACP concluding that the protein exists in equilibrium between two states: a compact and an extended conformation.