Investigation into the membrane interaction of E.coli Pencillin binding protein 4

Abstract

The low molecular mass penicillin-binding proteins (PBPs) of E. coli play a regulatory role in the terminal stages of peptidoglycan biosynthesis. PBPs 4, 5 and 6 have been shown to possess DD-carboxypeptidase activity with PBP4 also having an endopeptidase activity. PBP5 and 6 are proposed to anchor to the periplasmic face of the inner membrane via C-terminal ahelices whereas the association of PBP4 with the membrane is more ambiguous and may involve a specific binding site and/or C-terminal amphiphilic a-helical interaction. PBP4 has also been tentatively proposed to function as part of an enzyme complex that serves to insert new peptidoglycan strands into the pre-existing layer surrounding the bacterial
cell.
The effect that P4, a C-terminal homologue of PBP4, had on an artificial bilayer has been investigated using 2H NMR. This has been compared with the interactions of P5, a PBP5 C-terminal homologue, and melittin, a well documented membrane interactive protein of similar molecular mass. In the NMR experiments it was found that splitting reduced with increasing
P4 concentration (0 - 8.27 mg.g5, P<0.088. A similar, though less convincing, effect was observed with P5 (0 - 6.94 mg.gj, P<0.86, and melittin (0 - 3.06 mg.g 1 ), Pc0.39. This data implies that P4 interferes with the lamellar structure to a greater degree than either PS or melittin, possibly causing disruption, and would support a role for the C-terminus of PBP4 in membrane binding.
In monolayer experiments, P4 produced an increase in surface pressure of between 1 and 7.5 mN m 1 , with monolayers formed with either dioleoylphosphatodylglycerol (DOPO), dioleoylphosphatidylethanolamine (DOPE), a mixed lipid system (75% DOPE, 20% DOPG, 5% cardiolipin) or E. coli lipid extract. Similar pressure increases were observed with the addition of NaC1 (1 and 10 gM) to the sub phase of DOPG and lipid extract monolayers. This data confirms that the C-terminus of the protein would be able to interact with the bilayer interface and implies that the driving force behind the association is hydrophobic in nature rather than electrostatic. These pressure changes were higher than those caused by the whole protein (PBP4.his) which was found to interact better with ionic DOPO monolayers, producing an increase in pressure of 3 mN m 4 , compared with no increase in pressure for the zwitterionic DOPE monolayers. PBP4.his also interacted with the E. coli lipid extract (2 mN rn4 ) better than with the mixed lipid monolayers (0.5 mN m 1 ). The addition of NaC1 (1 .tM) to the sub phase of DOPO and lipid extract monolayers prevented any interaction, indicating an electrostatic component to the membrane interaction mechanism of PBP4.
The combination of data from the monolayer and NMR experiments shows that whilst the C-terminus could contribute to membrane association there may be stabilising electrostatic interactions between the ectornembraneous domain and the lipid.
Cross-linking studies were also performed to investigate the possibility that PBP4 can interact with other membrane proteins. PBP4 was found to cross-link to another membrane protein implying that there is potential for PBP4 to bind to the membrane via a specific site as well as the potential to form an enzyme complex.
In conclusion, these results supports the theory of a specific binding site involving protein-protein interaction for PBP4 with a requirement for ionic lipids to stabilise the membrane interaction of the whole protein. The monolayer and NMR data would support a role for the C-terminus possibly in stabilising initial interactions at the periplasmic face of the membrane.