Folate biosynthesis in higher plants. cDNA cloning, heterologous expression, and characterization of dihydroneopterin aldolases

Abstract

Dihydroneopterin aldolase (EC 4.1.2.25) is one of the enzymes of folate synthesis that remains to be cloned and characterized from plants. This enzyme catalyzes conversion of 7,8-dihydroneopterin (DHN) to 6-hydroxymethyl-7,8-dihydropterin, and is encoded by the folB gene in Escherichia coli. The E. coli FolB protein also mediates epimerization of DHN to 7,8-dihydromonapterin. Searches of the Arabidopsis genome detected three genes encoding substantially diverged FolB homologs (AtFolB1-3, sharing 57%-73% identity), for which cDNAs were isolated. A fourth cDNA specifying a FolB-like protein (LeFolB1) was obtained from tomato (Lycopersicon esculentum) by reverse transcription-PCR. When overproduced in E. coli, recombinant AtFolB1, AtFolB2, and LeFolB1 proteins all had both dihydroneopterin aldolase and epimerase activities, and carried out the aldol cleavage reaction on the epimerization product, 7,8-dihydromonapterin, as well as on DHN. AtFolB3, however, could not be expressed in active form. Size exclusion chromatography indicated that the plant enzyme is an octamer, like the bacterial enzyme. Quantifying expression of the Arabidopsis genes by real-time reverse transcription-PCR showed that AtFolB1 and AtFolB2 messages occur at low levels throughout the plant, whereas the AtFolB3 mRNA was detected only in siliques and only with an extremely low abundance. Sequence comparisons and phylogenetic analysis of FolB homologs from 16 plants indicated that their N-terminal regions are highly variable, and that most species have a small number of FolB genes that diverged after separation of the lineages leading to families. The substantial divergence of FolB homologs in Arabidopsis and other plants suggests that some of them may act on substrates other than DHN.

Figure 1.

The pteridine branch of the folate synthesis pathway. The aldolase and epimerase reactions catalyzed by E. coli DHN aldolase (FolB) are boxed. The phosphate groups of DHN triphosphate are removed in two steps. It is unclear whether the first step (elimination of pyrophosphate) is chemical or enzymatic (Suzuki and Brown, 1974; De Saizieu et al., 1995); the second step can be mediated by various nonspecific phosphomonoesterases (Suzuki and Brown, 1974). DHN-TP, 7,8-dihydroneopterin triphosphate; DHN, 7,8-dihydroneopterin; DHM, 7,8-dihydromonapterin; HMDHP, 6-hydroxymethyl-7,8-dihydropterin.

Figure 2.

Alignment of the deduced amino acid sequences of Arabidopsis and tomato FolB homologs with the DHN aldolases of E. coli (FolB), S. aureus, and Synechocystis. Identical residues are shaded in black, similar residues in gray. Dashes are gaps introduced to maximize alignment. Underlining indicates an additional exon present in the AtFolB1 and AtFolB3 sequences. The arrow marks the position of an intron common to all three Arabidopsis sequences. Asterisks indicate conserved residues implicated in catalysis or pteridine binding in the S. aureus enzyme (Hennig et al., 1998). EcFolB, E. coli FolB (GenBank accession number P31055); SaDHNA, S. aureus DHN aldolase (GenBank P56740); SsDHNA, Synechocystis sp. PCC 6803 DHN aldolase (GenBank S76177); AtFolB1, gene At3g11750; AtFolB2, gene At5g62980; AtFolB3, gene At3g21730; LeFolB1, tomato DHN aldolase (GenBank AY422466).

Figure 3.

HPLC-fluorometric analysis of the products formed from DHN by recombinant AtFolB1-ΔE protein. A, Standard pteridines. Retention times: DHN, 4.2 min; DHM, 5.7 min; HMDHP, 12.8 min. B, Reaction mixture in which AtFolB1-ΔE protein was incubated with DHN for 2.5 min. Samples were treated with acidic I₂/KI before HPLC separation to convert pteridines to their fully oxidized, fluorescent forms.

Figure 4.

Quantification of AtFolB1, AtFolB2, and AtFolB3 mRNAs in Arabidopsis organs. Levels of mRNA were determined by real-time quantitative RT-PCR, using primers spanning two exons of each gene. Note that values for AtFolB3 are multiplied by 1,000. Roots were from hydroponically grown plants. Three independent RNA extracts were made of each organ, and triplicate mRNA determinations were made on each extract. Data are means of all nine determinations ±SE. Internal RNA standards were used to estimate recovery from RT-PCR of each sample; the data shown are corrected for recovery.

Figure 5.

Unrooted phylogram generated using the neighbor-joining method and 1,000 bootstrap reiterations of 30 FolB-like sequences from 16 angiosperm species. Colored ovals show members of the Brassicaceae, Fabaceae, and Gramineae. Asterisks show branchpoints with bootstrap values >70%. The arrow indicates where the tree is rooted when prokaryotic FolB sequences are included as an outgroup in the analysis. Key to species: At, Arabidopsis thaliana; Bn, Brassica napus; Gm, Glycine max; Mt, Medicago truncatula; Lj, Lotus japonicus; Ls, Lactuca sativa; Os, Oryza sativa; Ta, Triticum aestivum; Hv, Hordeum vulgare; Zm, Zea mays; Px, Populus balsamifera × Populus deltoides; Vv, Vitis vinifera; St, Solanum tuberosum; Le, Lycopersicon esculentum; Bv, Beta vulgaris; Ga, Gossypium arboreum. The three sequences for rice (O. sativa) are based on genome data and at least one EST in each case. Sequences for other species are deduced from ESTs; all apparently encode complete proteins except for L. japonicus, where both sequences lack a few C-terminal residues.