Substrate flexibility of Mycoplasma fermentans mf1 phosphorylcholine transferase

Abstract

Zwitterionic modifications of glycans such as phosphorylcholine or phosphoethanolamine occur in a wide range of prokaryotic and eukaryotic organisms and are known for interaction with the mammalian immune system. Unlike the biosynthesis of membrane phospholipids which is well elucidated, very little is known about the transfer of zwitterionic phosphodiester moieties onto glycoconjugates. The presence and function of relevant enzymes has been suggested by gene knockout or mutation and corresponding aberrant phosphorylcholine metabolism. In the current study, the Mycoplasma fermentans phosphorylcholine transferase mf1, with previously confirmed in-vitro activity synthesizing phosphorylcholine-α-glucosyl-1,2-dipalmitoyl glycerol, is demonstrated to not only transfer phosphorylcholine but also phosphoethanolamine from CDP-ethanolamine. Moreover, mf1 is capable of using the β-configuration of the presumed natural substrate but transfers neither to simpler substrates with glucose moieties such as β-D-octyl-glucopyranoside nor to an extended lipid substrate with an additional galactose residue. These findings suggest a certain, but limited, substrate flexibility for bacterial PC-transferases. Mf1 activity is inhibited by β-glycerophosphate, an isomer of part of CDP-glycerol which is known to compete with CDP-ribitol in enzymatic reactions catalyzed by fukutin, a human protein sharing structural homology with mf1. For the first time, a phosphorylcholine transferase, mf1, could be biochemically characterized in vitro and its lipid products with zwitterionic phosphodiesters attached could be detected specifically with the pentraxin serum amyloid P.

Keywords: Mycoplasma fermentans; LicD; PC-transferase; Phosphocholine; Phosphorylcholine.

Fig. 1

A: Chemical structure of both lipid substrates synthesized for enzymatic activity and characterization experiments (upper structure: α-anomer, lower structure: β-anomer). B: Western Blot of the BL21-AI E. coli cell lysate using an anti-His antibody for detection. A distinct band at approximately 31 kD is indicated with an arrow. C: Putative reaction catalyzed by mf1 resulting in phosphorylcholine-glucopyranosyl-1,2-dipalmitoylglycerol and cytidine monophosphate (CMP). D: Structural model of mf1 predicted by AlphaFold 3. Protein sequence search suggested two models, cholinephosphotransferase mf1 (red) and an uncharacterized protein of Mycoplasma fermentans PG18 (green). Both models were superimposed and depicted using Schrödinger PyMOL (RMSD 0.454). C-terminus is shown on the left, region of beta sheets surrounded by alpha helices on the right

Fig. 2

Transferase activity of mf1 in E. coli BL21-AI cell lysate. Assays were incubated for 16 h at 37 °C and samples measured by MALDI-TOF MS and MSMS. A and B: Phosphorylcholine transferase activity assay using α- or β-glucosyldipalmitoyl glycerol as substrate, respectively (m/z 753 corresponds to the lipid substrate, sodium adduct; successful transfer of phosphorylcholine indicated as m/z 896; H⁺). C and D: Phosphoethanolamine transferase activity assay using α-glucosyldipalmitoyl glycerol as substrate. Samples were measured in positive and negative ion mode MS (successful transfer indicated as m/z 876; sodium adduct, in positive and m/z 852 in negative MS corresponding to phosphoethanolamine-α-glucosyldipalmitoyl glycerol). E and F: MSMS of m/z 896 phosphorylcholine-α- and β-glucosyldipalmitoyl glycerol, respectively. Inverse ratios of signal intensities of fragment ions at m/z 184 (PC-related, PC + H₂O) and 328 (HexPC) depending on the configuration of the substrate were observed. G: Negative ion mode MSMS of m/z 852, phosphoethanolamine-α-glucosyldipalmitoyl glycerol, whereby the fragment ions at m/z 140 & 284 are PE + H₂O or HexPE. H: Serum amyloid protein (SAP) dot blot of mf1 assays. Assays were performed and incubated as described, before 1 µL was loaded on nitrocellulose membrane. I and IV: α- and β-glucosyldipalmitoyl glycerol containing assay, respectively, without enzymatic treatment. II: α-configuration of substrate + CDP-choline after enzymatic treatment with mf1. III: α-configuration of substrate + CDP-ethanolamine after enzymatic treatment with mf1. V: β-configuration of substrate + CDP-choline after enzymatic treatment with mf1

Fig. 3

Removal of phosphorylcholine by in-house expressed PDase with similar reaction mechanism to an enzyme described [18]. Assays were first incubated with mf1, phosphorylcholine-product formation confirmed and afterwards treated with PDase, incubated at 37 °C for 16 h and samples analyzed by MALDI-TOF MS. A-D: Spectra of α-glucosyldipalmitoyl glycerol and β-glucosyldipalmitoyl glycerol post mf1 phosphorylcholine transfer before and after PDase addition, respectively. Efficient removal of phosphorylcholine from the glucose moiety is visible as decline or absence of the signal at m/z 896 after PDase addition. E and F: Activity assay of mf1 on the glucosyl-substrates extended by one galactose residue. α- and β-glucosyldipalmitoyl glycerol substrates were treated with β-1,4-galactosyltransferase from bovine milk for 16 h at 37 °C before mf1 was added to the reaction. Successful transfer of galactose visible as m/z 915 (sodium adduct); activity of mf1 detected exclusively on the remaining substrate without galactose (product at m/z 896 as in Fig. 2). Inverse ratios of product formation, galacto-glucosyldipalmitoyl glycerol and phosphorylcholine-glucosyldipalmitoyl glycerol for α- and β- substrate configurations are observed

Fig. 4

Effects of temperature and different metal ions on mf1 activity. Assays using both, the α- (circles in A, black color in B) and the β-configuration (triangles in A, grey color in B) of the substrate were performed as described and incubated for 16 h. Samples were measured by MALDI-TOF MS

Fig. 5

Phosphorylcholine transferase activity of mf1 followed over time. Samples were analyzed by MALDI-TOF MS at different time points and the ratio of product to substrate calculated. Assays were incubated at 37 °C using both, the α- (black) and β-configuration (grey) of the substrate

Fig. 6

Effects of pH and buffer salts on mf1 activity. A: Assays using only the α-configuration of the substrate were performed in AMPD buffer and incubated for 16 h. Samples were measured by MALDI-TOF MS. B: Both, the α- (black) and the β-configuration (gray) of the substrate were used