Probing lipidation of membrane active peptides and integral membrane proteins by liquid chromatography-mass spectrometry and ion mobility separation-mass spectrometry

Abstract

Liquid chromatography coupled with mass spectrometry (LC-MS) and tandem mass spectrometry (LC-MS2) are shown to have the sensitivity and functionality to detect protein/peptide modifications by fatty acyl chains in vitro and in vivo studies. Further analysis was also performed by direct infusion ion mobility separation coupled with mass spectrometry (IMS-MS) or tandem mass spectrometry (IMS-MS2).
Peptide lipidation in vitro was investigated using the membrane active peptide, melittin. Non-enzymatic melittin lipidation by lysophospholipids has been observed for the first time. When the effect of lysophospholipids was studied in direct competition with diacylphospholipids, the acyl transfer from the lysophospholipids is seen to be preferential with acylation visible after just 3 hour. The longer the interaction time, the greater the amount and number of modifications with double and triple acylation observed after 96 hour. The locations of the modifications identified through LC-MS2 were assigned on different sites of the peptide, including N-terminus, K7, S18, K21, K23, R22 and R24 and with the highest reactivity towards N-terminus and K23. Comparing the lipidation of synthetic melittin (SynM) with the lipidation of naturally occurring melittin from venom of honey bee (BVM) highlights the effect of the PLA2 enzyme that is naturally present in BVM. Here, the action of the enzyme to hydrolyse the diacylphospholipid at the sn-2 position to give the corresponding lysophospholipid is reflected in the acyl transfer to the BVM such that the resulting lysophospholipid clearly dominates the acyl transfer to BVM. In contrast, the acyl transfer to SynM clearly demonstrates that acyl transfer is possible in the absence of an enzyme.
In vivo protein lipidation of one of the most abundant integral membrane proteins in mammalian eye lens, AQP0, was also studied. A wide range of acyl groups are shown to modify this protein at the known modification sites, N-terminus and at the amino acid residue K238, many of which are reported here for the first time. These acyl group modifications reflect the biological lipid composition of the membrane leaflet that the acylation sites are proximal to.
In an attempt to further distinguish between different forms of lipidated melittin, whether with the same acyl chain modification to different amino acids or to discriminate between palmitoylation and oleoylation modifications, travelling wave ion mobility spectrometry (TWIMS) coupled with MS or MS2 was applied. Results suggested that resolving positional isomers of diacylphospholipids and lysophospholipids (sn-1 vs sn-2 positions) is not possible under the conditions described herein. However, the presence of fatty acyl chains covalently bound to melittin change the conformation of acylated melittin in the gas-phase such that for lower charge states it is possible to suggest a small degree of separation between palmitoylated and oleoylated melittin or their isomers including acylation on N-terminus vs K23. This small degree of separation is enough so that when combined with tandem mass spectrometry, the time-aligned product ion spectra are clearer and improve characterisation.
To conclude, the research in this thesis has shown that two of the most abundant biomolecules, lipid and peptides/proteins that are known to exist in close proximity to each other, or interact with each other, are not as chemically inert as previously thought. This reactivity has been reflected herein via aminolysis reaction between membrane lipid composition and each of membrane active peptide, melittin and integral membrane protein, AQP0.