Biosynthesis of phosphoserine in the Methanococcales

Abstract

Methanococcus maripaludis and Methanocaldococcus jannaschii produce cysteine for protein synthesis using a tRNA-dependent pathway. These methanogens charge tRNA(Cys) with l-phosphoserine, which is also an intermediate in the predicted pathways for serine and cystathionine biosynthesis. To establish the mode of phosphoserine production in Methanococcales, cell extracts of M. maripaludis were shown to have phosphoglycerate dehydrogenase and phosphoserine aminotransferase activities. The heterologously expressed and purified phosphoglycerate dehydrogenase from M. maripaludis had enzymological properties similar to those of its bacterial homologs but was poorly inhibited by serine. While bacterial enzymes are inhibited by micromolar concentrations of serine bound to an allosteric site, the low sensitivity of the archaeal protein to serine is consistent with phosphoserine's position as a branch point in several pathways. A broad-specificity class V aspartate aminotransferase from M. jannaschii converted the phosphohydroxypyruvate product to phosphoserine. This enzyme catalyzed the transamination of aspartate, glutamate, phosphoserine, alanine, and cysteate. The M. maripaludis homolog complemented a serC mutation in the Escherichia coli phosphoserine aminotransferase. All methanogenic archaea apparently share this pathway, providing sufficient phosphoserine for the tRNA-dependent cysteine biosynthetic pathway.

FIG. 1.

Biosynthesis of l-phosphoserine in the Methanococcales. PGDH catalyzes the oxidation of d-3-phosphoglycerate to produce phosphohydroxypyruvate, concomitant with the reduction of NAD⁺. A broad-specificity aminotransferase (AspAT) catalyzes the transamination reaction from l-glutamate to produce l-3-phosphoserine and α-ketoglutarate. Phosphoserine phosphatase produces l-serine for protein synthesis and glycine production. In a reaction with homocysteine (Hcy), phosphoserine produces cystathionine (50). Alternatively, Sep can be used to aminoacylate tRNA^Cys, which can be converted to Cys-tRNA^Cys by a sulfide transferase enzyme (38).

FIG. 2.

Cell extract of M. maripaludis catalyzes the transamination of aspartate and phosphohydroxypyruvate to produce phosphoserine. Reaction mixtures containing 40 μM PHP, 5 mM l-aspartate, and 44 μg M. maripaludis cell extract were incubated for 30 min at 30°C. The CBI derivatives of the amino acids were separated by reversed-phase HPLC and detected by their fluorescence. All of the PHP in the reaction mix was converted to phosphoserine, based on comparison of the phosphoserine-CBI peak area to an external standard. The aspartate-CBI peak is off scale in these overlaid chromatograms.

FIG. 3.

His₆-MMP1588 (PGDH) and His₁₀-MJ0959 (AspAT) were purified to homogeneity by affinity chromatography. Proteins were separated by SDS-PAGE and stained with Coomassie blue dye. Lanes M, broad-range protein standards (Bio-Rad), with masses indicated in kDa; lane 1, purified His₆-MMP1588; lane 2, purified His₁₀-MJ0959.

FIG. 4.

The M. maripaludis MMP0391 gene complements the serC mutation of E. coli KL285. (A) M9 minimal medium with glucose containing all 20 amino acids supports the growth of E. coli KL285 containing (clockwise from top) pKQV4 (E. coli serC), empty pKQV4 vector, and pKQV4 (MMP0391). (B) Minimal medium containing 19 amino acids (without serine). (C) Minimal medium containing 19 amino acids (without serine) and 1 mM IPTG. All media contain 100 μg ml⁻¹ ampicillin.

FIG. 5.

Purified MJ0959 protein catalyzes the transamination of aspartate and phosphohydroxypyruvate to produce phosphoserine and alanine. Reaction mixtures were incubated at 50°C. Fluorescent CBI derivatives were separated and analyzed by HPLC as described in the legend to Fig. 2.

FIG. 6.

Phylogeny of class V aspartate aminotransferase homologs constructed using a Bayesian inference method. The branches are labeled with organism names and the GenBank or SwissProt accession number for each sequence. The tree is arbitrarily rooted. Bar, 0.1 amino acid change expected per site. Numbers near each interior branch are clade credibility values, or the fraction of trees that contain each cluster of sequences shown in the consensus tree. This tree is generally congruent with trees produced using protein maximum likelihood and neighbor-joining distance methods. Branches with high clade credibility values also have high bootstrap values (from 100 neighbor-joining distance trees). Previously characterized enzymes are labeled AEPT (2-aminoethylphosphonate transaminase), SPT (serine-pyruvate transaminase), AGAT (alanine-glyoxylate aminotransferase), and SGAT (serine-glyoxylate aminotransferase).