BackgroundFolate synthesis and salvage pathways are relatively well known from classical biochemistry and genetics but they have not been subjected to comparative genomic analysis. The availability of genome sequences from hundreds of diverse bacteria, and from Arabidopsis thaliana, enabled such an analysis using the SEED database and its tools. This study reports the results of the analysis and integrates them with new and existing experimental data.ResultsBased on sequence similarity and the clustering, fusion, and phylogenetic distribution of genes, several functional predictions emerged from this analysis. For bacteria, these included the existence of novel GTP cyclohydrolase I and folylpolyglutamate synthase gene families, and of a trifunctional p-aminobenzoate synthesis gene. For plants and bacteria, the predictions comprised the identities of a 'missing' folate synthesis gene (folQ) and of a folate transporter, and the absence from plants of a folate salvage enzyme. Genetic and biochemical tests bore out these predictions.ConclusionFor bacteria, these results demonstrate that much can be learnt from comparative genomics, even for well-explored primary metabolic pathways. For plants, the findings particularly illustrate the potential for rapid functional assignment of unknown genes that have prokaryotic homologs, by analyzing which genes are associated with the latter. More generally, our data indicate how combined genomic analysis of both plants and prokaryotes can be more powerful than isolated examination of either group alone.

Plants are the main source of folate in human diets, but many fruits, tubers, and seeds are poor in this vitamin, and folate deficiency is a worldwide problem. Plants synthesize folate from pteridine, p-aminobenzoate (PABA), and glutamate moieties. Pteridine synthesis capacity is known to drop in ripening tomato fruit; therefore, we countered this decline by fruit-specific overexpression of GTP cyclohydrolase I, the first enzyme of pteridine synthesis. We used a synthetic gene based on mammalian GTP cyclohydrolase I, because this enzyme is predicted to escape feedback control in planta. This engineering maneuver raised fruit pteridine content by 3-to 140-fold and fruit folate content by an average of 2-fold among 12 independent transformants, relative to vector-alone controls. Most of the folate increase was contributed by 5-methyltetrahydrofolate polyglutamates and 5,10-methenyltetrahydrofolate polyglutamates, which were also major forms of folate in control fruit. The accumulated pteridines included neopterin, monapterin, and hydroxymethylpterin; their reduced forms, which are folate biosynthesis intermediates; and pteridine glycosides not previously found in plants. Engineered fruit with intermediate levels of pteridine overproduction attained the highest folate levels. PABA pools were severely depleted in engineered fruit that were high in folate, and supplying such fruit with PABA by means of the fruit stalk increased their folate content by up to 10-fold. These results demonstrate that engineering a moderate increase in pteridine production can significantly enhance the folate content in food plants and that boosting the PABA supply can produce further gains.

GTP cyclohydrolase I (GCHI) mediates the first and committing step of the pterin branch of the folate-synthesis pathway. In microorganisms and mammals, GCHI is a homodecamer of Ϸ26-kDa subunits. Genomic approaches identified tomato and Arabidopsis cDNAs specifying Ϸ50-kDa proteins containing two GCHI-like domains in tandem and indicated that such bimodular proteins occur in other plants. Neither domain of these proteins has a full set of the residues involved in substrate binding and catalysis in other GCHIs. The tomato and Arabidopsis cDNAs nevertheless encode functional enzymes, as shown by complementation of a yeast fol2 mutant and by assaying GCHI activity in extracts of complemented yeast cells. Neither domain expressed separately had GCHI activity. Recombinant tomato GCHI formed dihydroneopterin triphosphate as reaction product, as do other GCHIs, but unlike these enzymes it did not show cooperative behavior and was inhibited by its substrate. Denaturing gel electrophoresis verified that the bimodular GCHI polypeptide is not cleaved in vivo into its component domains, and size-exclusion chromatography indicated that the active enzyme is a dimer. The deduced tomato and Arabidopsis GCHI polypeptides lack overt targeting sequences and thus are presumably cytosolic, in contrast to other plant folate-synthesis enzymes, which are mitochondrial proteins with typical signal peptides. GCHI mRNA and protein are strongly in expressed unripe tomato fruits, implying that fruit folate is made in situ rather than imported. As ripening advances, GCHI expression declines sharply, and folate content drops, suggesting that folate synthesis fails to keep pace with turnover.T etrahydrofolate and its derivatives (folates) are essential cofactors for one-carbon transfer reactions in all organisms. Similar to bacteria and yeasts, plants make folates de novo from pterin, p-aminobenzoate (PABA), and glutamate moieties (1, 2). In contrast, humans and other mammals lack a complete folatesynthesis pathway and thus need dietary folate. Because plant foods are major folate sources, and folate deficiency is a global health problem, enhancing plant folate content is a prime target for metabolic engineering (1, 3). This engineering demands knowledge of the biosynthetic pathway.The plant folate-synthesis pathway is not understood fully, but is most probably similar to that in bacteria (1, 2). The pterin hydroxymethyldihydropteroate is formed from GTP, and PABA from chorismate. The pterin and PABA units are condensed, glutamylated, and reduced to give tetrahydrofolate, and a polyglutamyl tail is added (4). Plant genes and enzymes for the final five steps have been characterized (5, 6), and the enzymes have been shown to be mitochondrial (7,8). Far less is known for plants about the early steps that produce the pterin and PABA moieties.The first step of pterin synthesis is of special interest, because it commits GTP to pterin production and is considered to control flux into the pathway (9, 10). This step, mediated by GTP cyclohydrolase I (GC...

5-Formyltetrahydrofolate (5-CHO-THF) is formed viaa second catalytic activity of serine hydroxymethyltransferase (SHMT) and strongly inhibits SHMT and other folate-dependent enzymes in vitro. The only enzyme known to metabolize 5-CHO-THF is 5-CHO-THF cycloligase (5-FCL), which catalyzes its conversion to 5,10-methenyltetrahydrofolate. Because 5-FCL is mitochondrial in plants and mitochondrial SHMT is central to photorespiration, we examined the impact of an insertional mutation in the Arabidopsis 5-FCL gene (At5g13050) under photorespiratory (30 and 370 mol of CO 2 mol ؊1 ) and non-photorespiratory (3200 mol of CO 2 mol ؊1 ) conditions. The mutation had only mild visible effects at 370 mol of CO 2 mol ؊1 , reducing growth rate by ϳ20% and delaying flowering by 1 week. However, the mutation doubled leaf 5-CHO-THF level under all conditions and, under photorespiratory conditions, quadrupled the pool of 10-formyl-/5,10-methenyltetrahydrofolates (which could not be distinguished analytically). At 370 mol of CO 2 mol ؊1 , the mitochondrial 5-CHO-THF pool was 8-fold larger in the mutant and contained most of the 5-CHO-THF in the leaf. In contrast, the buildup of 10-formyl-/5,10-methenyltetrahydrofolates was extramitochondrial. In photorespiratory conditions, leaf glycine levels were up to 46-fold higher in the mutant than in the wild type. Furthermore, when leaves were supplied with 5-CHO-THF, glycine accumulated in both wild type and mutant. These data establish that 5-CHO-THF can inhibit SHMT in vivo and thereby influence glycine pool size. However, the near-normal growth of the mutant shows that even exceptionally high 5-CHO-THF levels do not much affect fluxes through SHMT or any other folate-dependent reaction, i.e. that 5-CHO-THF is well tolerated in plants. 5-Formyltetrahydrofolate (5-CHO-THF)1 is formed from 5,10-methenyltetrahydrofolate (5,10-CHϭTHF) by a hydrolytic reaction catalyzed by serine hydroxymethyltransferase (SHMT) in the presence of glycine (1, 2). Spontaneous chemical hydrolysis of 5,10-CHϭTHF may be a minor additional source (3). 5-CHO-THF is the most stable natural folate and the most enigmatic, for it is the only one that does not serve as a cofactor in one-carbon metabolism. Instead, 5-CHO-THF is a potent inhibitor of SHMT and most other folate-dependent enzymes in vitro (4, 5). 5-CHO-THF probably acts as a stable storage form of folate in seeds and fungal spores (5-7), but it is not clear what role, if any, it plays in metabolically active tissues (8). This question is particularly pertinent for leaves. Leaf mitochondria have very high levels of SHMT and, during photorespiration, receive a massive influx of glycine (which leads to a matching SHMT-mediated glycine 3 serine flux) (9). Conditions in leaf mitochondria therefore favor 5-CHO-THF formation (Fig. 1). Indeed, 5-CHO-THF can comprise 50% of the folate pool in leaf mitochondria (10, 11), which is far more than in mammalian mitochondria (12)(13)(14). Furthermore, 5-CHO-THF is reported to make up 14 -40% of the folate pool in leaves and o...

Cyanobacterial and plant genomes encode proteins with some similarity to the folate and biopterin transporters of the trypanosomatid parasite Leishmania. The Synechocystis slr0642 gene product and its closest Arabidopsis homolog, the At2g32040 gene product, are representative examples. Both have 12 probable transmembrane domains, and the At2g32040 protein has a predicted chloroplast transit peptide. When expressed in Escherichia coli pabA pabB or folE, mutants, which are unable to produce or take up folates, the slr0642 protein and a modified At2g32040 protein (truncated and fused to the N terminus of slr0642) enabled growth on 5-formyltetrahydrofolate or folic acid but not on 5-formyltetrahydrofolate triglutamate, demonstrating that both proteins mediate folate monoglutamate transport. Both proteins also mediate transport of the antifolate analogs methotrexate and aminopterin, as evidenced by their ability to greatly increase the sensitivity of E. coli to these inhibitors. The full-length At2g32040 polypeptide was translocated into isolated pea chloroplasts and, when fused to green fluorescent protein, directed the passenger protein to the envelope of Arabidopsis chloroplasts in transient expression experiments. At2g32040 transcripts were present at similar levels in roots and aerial organs, indicating that the protein occurs in non-green plastids as well as chloroplasts. Insertional inactivation of At2g32040 significantly raised the total folate content of chloroplasts and lowered the proportion of 5-methyltetrahydrofolate but did not discernibly affect growth. These findings establish conservation of function among folate and biopterin transporter family proteins from three kingdoms of life.

SummaryFolates in vivo undergo oxidative cleavage, giving pterin and p-aminobenzoylglutamate (pABAGlu) moieties. These breakdown products are excreted in animals, but their fate is unclear in microorganisms and unknown in plants. As indirect evidence from this and previous studies strongly suggests that plants can have high folatebreakdown rates (approximately 10% per day), salvage of the cleavage products seems likely. Four sets of observations support this possibility. First, cleavage products do not normally accumulate: pools of pABAGlu (including its polyglutamyl forms) are equivalent to, at most, 4-14% of typical total folate pools in Arabidopsis thaliana, Lycopersicon esculentum and Pisum sativum tissues. Pools of the pterin oxidation end-product pterin-6-carboxylate are, likewise, fairly small (3-37%) relative to total folate pools. Second, little pABAGlu built up in A. thaliana plantlets when net folate breakdown was induced by blocking folate synthesis with sulfanilamide. Third, A. thaliana and L. esculentum tissues readily converted supplied breakdown products to folate synthesis precursors: pABAGlu was hydrolysed to p-aminobenzoate and glutamate, and dihydropterin-6-aldehyde was reduced to 6-hydroxymethyldihydropterin. Fourth, both these reactions were detected in vitro; the reduction used NADPH as cofactor. An alternative salvage route for pABAGlu, direct reincorporation into dihydrofolate via the action of dihydropteroate synthase, appears implausible from the properties of this enzyme. We conclude that plants are excellent organisms in which to explore the biochemistry of folate salvage.

␥-Glutamyl hydrolase (GGH, EC 3.4.19.9) catalyzes removal of the polyglutamyl tail from folyl and p-aminobenzoyl polyglutamates. Plants typically have one or a few GGH genes; Arabidopsis has three, tandemly arranged on chromosome 1, which encode proteins with predicted secretory pathway signal peptides. Two representative Arabidopsis GGH proteins, AtGGH1 and AtGGH2 (the At1g78660 and At1g78680 gene products, respectively) were expressed in truncated form in Escherichia coli and purified. Both enzymes were active as dimers, had low K m values (0.5-2 M) for folyl and p-aminobenzoyl pentaglutamates, and acted as endopeptidases. However, despite 80% sequence identity, they differed in that AtGGH1 cleaved pentaglutamates, mainly to di-and triglutamates, whereas AtGGH2 yielded mainly monoglutamates. Analysis of subcellular fractions of pea leaves and red beet roots established that GGH activity is confined to the vacuole and that this activity, if not so sequestered, would deglutamylate all cellular folylpolyglutamates within minutes. Purified pea leaf vacuoles contained an average of 20% of the total cellular folate compared with ϳ50 and ϳ10%, respectively, in mitochondria and chloroplasts. The main vacuolar folate was 5-methyltetrahydrofolate, of which 51% was polyglutamylated. In contrast, the principal mitochondrial and chloroplastic forms were 5-formyl-and 5,10-methenyltetrahydrofolate polyglutamates, respectively. In beet roots, 16 -60% of the folate was vacuolar and was again mainly 5-methyltetrahydrofolate, of which 76% was polyglutamylated. These data point to a hitherto unsuspected role for vacuoles in folate storage. Furthermore, the paradoxical co-occurrence of GGH and folylpolyglutamates in vacuoles implies that the polyglutamates are somehow protected from GGH attack.

Methylenetetrahydrofolate reductase (MTHFR; EC 1.5.1.20) is the sole enzyme responsible for generation of 5-methyltetrahydrofolate, which is required for methionine synthesis and provision of methyl groups via S-adenosylmethionine. Genome analysis showed that Leishmania species, unlike Trypanosoma brucei and Trypanosoma cruzi, contain genes encoding MTHFR and two distinct methionine synthases. Leishmania MTHFR differed from those in other eukaryotes by the absence of a C-terminal regulatory domain. L. major MTHFR was expressed in yeast and recombinant enzyme was produced in Escherichia coli. MTHFR was not inhibited by S-adenosylmethionine and, uniquely among folate-metabolizing enzymes, showed dual-cofactor specificity with NADH and NADPH under physiological conditions. MTHFR null mutants (mthfr ؊ ) lacked 5-methyltetrahydrofolate, the most abundant intracellular folate, and could not utilize exogenous homocysteine for growth. Under conditions of methionine limitation mthfr ؊ mutant cells grew poorly, whereas their growth was normal in standard culture media. Neither in vitro MTHFR activity nor the growth of mthfr ؊ mutants or MTHFR overexpressors were differentially affected by antifolates known to inhibit parasite growth via targets beyond dihydrofolate reductase and pteridine reductase 1. In a mouse model of infection mthfr ؊ mutants showed good infectivity and virulence, indicating that sufficient methionine is available within the parasitophorous vacuole to meet the needs of the parasite.