O-Phospho-D-serine

To test whether cellular protein kinases exist that phosphorylate D-amino acid residues, a method was developed for separating O-phospho-D-serine from O-phospho-L-serine and O-phospho-L-tyrosine from O-phospho-D-tyrosine. This was accomplished by converting these amino acids to the L-leucyl dipeptide derivatives followed by separation of the diastereomers by anion-exchange high-performance liquid chromatography. The enantiomeric content of these D- and L-residues were measured in hydrolysates of 32P-labeled proteins produced by the protein kinases of human erythrocytes and the tyrosyl protein kinase of the Abelson leukemia virus. We found no measurable D-phosphoserine in erythrocyte membrane proteins under conditions where a 1% content of this residue relative to L-phosphoserine would have been detected. These values can be used to place an upper hypothetical limit on the fraction of erythrocyte protein kinase activity that is specific for serine residues in the D-configuration. In separate experiments, we examined the specificity of the tyrosyl protein kinases. We found that all of the phosphotyrosine that we isolated from the erythrocyte band 3 NH2-terminal fragment and from the autophosphorylation of the Abelson virus tyrosyl kinase was in the L-configuration.

The enteric protozoan parasite Entamoeba histolytica is a unicellular eukaryote that possesses both phosphorylated and non-phosphorylated serine metabolic pathways. In the present study, we described enzymological and functional characterization of phosphoserine aminotransferase (PSAT) from E. histolytica. E. histolytica PSAT (EhPSAT) showed maximum activity for the forward reaction at basic pH, dissimilar to mammalian PSAT, which showed sharp neutral optimum pH. EhPSAT activity was significantly inhibited by substrate analogs, O-phospho-d-serine, O-phospho-l-threonine, and O-acetylserine, suggesting possible regulation of the amoebic PSAT by these metabolic intermediates. Fractionation of the whole parasite lysate and rEhPSAT by anion exchange chromatography verified that EhPSAT represents a dominant PSAT activity. EhPSAT showed a close kinship to PSAT from bacteroides based on amino acid alignment and phylogenetic analyses, suggesting that E. histolytica gained this gene from bacteroides by lateral gene transfer. Comparisons of kinetic properties of recombinant PSAT from E. histolytica and Arabidopsis thaliana showed that EhPSAT possesses significantly higher affinity toward glutamate than the A. thaliana counterpart, which may be explained by significant differences in the isoelectric point and the substitution of arginine, which is involved the binding to the gamma-carboxylate moiety of glutamate, in Escherichia coli PSAT, to serine or threonine in E. histolytica or A. thaliana PSAT, respectively. Heterologous expression of EhPSAT successfully rescued growth defect of a serine-auxotrophic E. coli strain KL282, where serC was deleted, confirming its in vivo role in serine biosynthesis. Together with our previous demonstration of phosphoglycerate dehydrogenase, the present study reinforces physiological significance of the phosphorylated pathway in amoeba.

Phosphatidylserine, a negatively charged lipid, is exposed on the platelet membrane following cell stimulation, correlating with the expression of factor VIII receptors. We have explored the importance of the negative electrostatic potential of phosphatidylserine vs chemical moieties of phosphatidylserine for specific membrane binding of factor VIII. Fluorescein-labeled factor VIII bound to membranes containing 15% phosphatidic acid, a negatively charged phospholipid, with low affinity compared to phosphatidylserine-containing membranes. Binding was not specific as it was inhibited by other proteins in plasma. Factor VIII bound to membranes containing 10% phosphatidylserine in spite of a varying net charge provided by 0-15% stearylamine, a positively charged lipid. The soluble phosphatidylserine moiety, O-phospho-L-serine, inhibited factor VIII binding to phosphatidylserine-containing membranes with a Ki of 20 mM, but the stereoisomer, O-phospho-D-serine, was 5-fold less effective. Furthermore, binding of factor VIII to membranes containing synthetic phosphatidyl-D-serine was 5-fold less than binding to membranes containing phosphatidyl-L-serine. Membranes containing synthetic phosphatidyl-L-homoserine, differing from phosphatidylserine by a single methylene, supported high-affinity binding, but it was not specific as factor VIII was displaced by other plasma proteins. O-Phospho-L-serine also inhibited the binding of factor VIII to platelet-derived microparticles with a Ki of 20 mM, and the stereoisomer was 4-fold less effective. These results indicate that membrane binding of factor VIII is mediated by a stereoselective recognition O-phospho-L-serine of phosphatidylserine and that negative electrostatic potential is of lesser importance.