Abstract
Phosphorylcholine (ChoP) is a charged molecule that, in humans, is the major recognition component of platelet activating factor (PAF) by its cognate receptor, platelet activating factor receptor (PAFR). ChoP-like molecules have been detected on the surfaces of numerous respiratory pathogens where they confer persistence related phenotypes mediated through an increase in adhesion. Pseudomonas aeruginosa is a Gram-negative opportunistic pathogen which is often associated with nosocomial pneumonia. Previously, in P. aeruginosa a ChoP-like modification was detected using ChoP-specific antibodies on surface exposed elongation factor-Tu (EF-Tu), a normally cytoplasmic protein involved in protein chain elongation. Here we identify a gene, eftM, which is required for the modification of EF-Tu. To our surprise, mass spectrometry analysis of modified EF-Tu revealed the modification was not ChoP but rather trimethyl-lysine at amino acid residue 5. This trimethyl-lysine acts as a structural mimic of ChoP and confers similar phenotypes to P. aeruginosa as ChoP does to other respiratory pathogens. Bioinformatic analysis of EftM showed amino acid sequence similarity to S-adenosylmethionine (SAM)-dependent methyltransferases. We confirmed biochemically that EftM is able to bind SAM and used it to directly methylate lysine 5 of EF-Tu both in vivo and in vitro. Mass spectrometry analysis of products from these reactions confirmed that EF-Tu is trimethylated at lysine 5. In vivo, P. aeruginosa methylates EF-Tu only at temperatures closer to ambient, 25°C, and not at body temperature, 37°C. We show that this temperature-dependent methylation phenotype is not due to differences in transcription of eftM. The methyltransferase activity of the laboratory P. aeruginosa strain PAO1 EftM is slightly higher at 25°C compared to 37°C, while the rate of degradation of EftM expressed in P. aeruginosa is significantly decreased at 25°C compared to 37°C allowing for a longer half-life of the protein. The difference in the rate of degradation suggested an increase in protein stability at 25°C and was confirmed by heat pre-treatment of EftM in vitro. Pre-incubation of EftM at 37°C abolished methyltransferase activity while methyltransferase activity was retained when the enzyme was pre-incubated at 25°C. In the absence of evidence for transcriptional regulation of this phenomenon, these results may suggest that the in vivo temperature-dependent phenotype is due to differences in the steady state levels of the EftM protein at different temperatures.
In this work, we have discovered a modification that mimics a known bacterial adhesin in respiratory pathogens. It is possible that the ChoP-like modification seen in other respiratory pathogens is actually trimethyl-lysine, the addition of which could be mediated by a methyltransferase similar to EftM. The identification and characterization of EftM could have broad implications for both basic and applied biology. These include the potential identification of a novel export pathway of a normally cytoplasmic protein and insight into the bacterial regulation of protein synthesis. Additionally, the enzyme responsible for the methylation of EF-Tu may be a good target for novel therapeutics to prevent the establishment of respiratory infections.
