Mutations in PROSC Disrupt Cellular Pyridoxal Phosphate Homeostasis and Cause Vitamin-B6-Dependent Epilepsy.

Pyridoxal 5'-phosphate (PLP), the active form of vitamin B6, functions as a cofactor in humans for more than 140 enzymes, many of which are involved in neurotransmitter synthesis and degradation. A deficiency of PLP can present, therefore, as seizures and other symptoms that are treatable with PLP and/or pyridoxine. Deficiency of PLP in the brain can be caused by inborn errors affecting B6 vitamer metabolism or by inactivation of PLP, which can occur when compounds accumulate as a result of inborn errors of other pathways or when small molecules are ingested. Whole-exome sequencing of two children from a consanguineous family with pyridoxine-dependent epilepsy revealed a homozygous nonsense mutation in proline synthetase co-transcribed homolog (bacterial), PROSC, which encodes a PLP-binding protein of hitherto unknown function. Subsequent sequencing of 29 unrelated indivduals with pyridoxine-responsive epilepsy identified four additional children with biallelic PROSC mutations. Pre-treatment cerebrospinal fluid samples showed low PLP concentrations and evidence of reduced activity of PLP-dependent enzymes. However, cultured fibroblasts showed excessive PLP accumulation. An E.coli mutant lacking the PROSC homolog (ΔYggS) is pyridoxine sensitive; complementation with human PROSC restored growth whereas hPROSC encoding p.Leu175Pro, p.Arg241Gln, and p.Ser78Ter did not. PLP, a highly reactive aldehyde, poses a problem for cells, which is how to supply enough PLP for apoenzymes while maintaining free PLP concentrations low enough to avoid unwanted reactions with other important cellular nucleophiles. Although the mechanism involved is not fully understood, our studies suggest that PROSC is involved in intracellular homeostatic regulation of PLP, supplying this cofactor to apoenzymes while minimizing any toxic side reactions.

Evidence that COG0325 proteins are involved in PLP homeostasis.

Pyridoxal 5'-phosphate (PLP) is an essential cofactor for nearly 60 Escherichia coli enzymes but is a highly reactive molecule that is toxic in its free form. How PLP levels are regulated and how PLP is delivered to target enzymes are still open questions. The COG0325 protein family belongs to the fold-type III class of PLP enzymes and binds PLP but has no known biochemical activity although it occurs in all kingdoms of life. Various pleiotropic phenotypes of the E. coli COG0325 (yggS) mutant have been reported, some of which were reproduced and extended in this study. Comparative genomic, genetic and metabolic analyses suggest that these phenotypes reflect an imbalance in PLP homeostasis. The E. coli yggS mutant accumulates the PLP precursor pyridoxine 5'-phosphate (PNP) and is sensitive to an excess of pyridoxine but not of pyridoxal. The pyridoxine toxicity phenotype is complemented by the expression of eukaryotic yggS orthologs. It is also suppressed by the presence of amino acids, specifically isoleucine, threonine and leucine, suggesting the PLP-dependent enzyme transaminase B (IlvE) is affected. These genetic results lay a foundation for future biochemical studies of the role of COG0325 proteins in PLP homeostasis.

Essentiality of threonylcarbamoyladenosine (t(6)A), a universal tRNA modification, in bacteria.

Threonylcarbamoyladenosine (t(6)A) is a modified nucleoside universally conserved in tRNAs in all three kingdoms of life. The recently discovered genes for t(6)A synthesis, including tsaC and tsaD, are essential in model prokaryotes but not essential in yeast. These genes had been identified as antibacterial targets even before their functions were known. However, the molecular basis for this prokaryotic-specific essentiality has remained a mystery. Here, we show that t(6)A is a strong positive determinant for aminoacylation of tRNA by bacterial-type but not by eukaryotic-type isoleucyl-tRNA synthetases and might also be a determinant for the essential enzyme tRNA(Ile)-lysidine synthetase. We confirm that t(6)A is essential in Escherichia coli and a survey of genome-wide essentiality studies shows that genes for t(6)A synthesis are essential in most prokaryotes. This essentiality phenotype is not universal in Bacteria as t(6)A is dispensable in Deinococcus radiodurans, Thermus thermophilus, Synechocystis PCC6803 and Streptococcus mutans. Proteomic analysis of t(6)A(-) D. radiodurans strains revealed an induction of the proteotoxic stress response and identified genes whose translation is most affected by the absence of t(6)A in tRNAs. Thus, although t(6)A is universally conserved in tRNAs, its role in translation might vary greatly between organisms.

Cross kingdom functional conservation of the core universally conserved threonylcarbamoyladenosine tRNA synthesis enzymes.

Threonylcarbamoyladenosine (t(6)A) is a universal modification located in the anticodon stem-loop of tRNAs. In yeast, both cytoplasmic and mitochondrial tRNAs are modified. The cytoplasmic t(6)A synthesis pathway was elucidated and requires Sua5p, Kae1p, and four other KEOPS complex proteins. Recent in vitro work suggested that the mitochondrial t(6)A machinery of Saccharomyces cerevisiae is composed of only two proteins, Sua5p and Qri7p, a member of the Kae1p/TsaD family (L. C. K. Wan et al., Nucleic Acids Res. 41:6332-6346, 2013, http://dx.doi.org/10.1093/nar/gkt322). Sua5p catalyzes the first step leading to the threonyl-carbamoyl-AMP intermediate (TC-AMP), while Qri7 transfers the threonyl-carbamoyl moiety from TC-AMP to tRNA to form t(6)A. Qri7p localizes to the mitochondria, but Sua5p was reported to be cytoplasmic. We show that Sua5p is targeted to both the cytoplasm and the mitochondria through the use of alternative start sites. The import of Sua5p into the mitochondria is required for this organelle to be functional, since the TC-AMP intermediate produced by Sua5p in the cytoplasm is not transported into the mitochondria in sufficient amounts. This minimal t(6)A pathway was characterized in vitro and, for the first time, in vivo by heterologous complementation studies in Escherichia coli. The data revealed a potential for TC-AMP channeling in the t(6)A pathway, as the coexpression of Qri7p and Sua5p is required to complement the essentiality of the E. coli tsaD mutant. Our results firmly established that Qri7p and Sua5p constitute the mitochondrial pathway for the biosynthesis of t(6)A and bring additional advancement in our understanding of the reaction mechanism.

The flavoprotein Mcap0476 (RlmFO) catalyzes m5U1939 modification in Mycoplasma capricolum 23S rRNA.

Efficient protein synthesis in all organisms requires the post-transcriptional methylation of specific ribosomal ribonucleic acid (rRNA) and transfer RNA (tRNA) nucleotides. The methylation reactions are almost invariably catalyzed by enzymes that use S-adenosylmethionine (AdoMet) as the methyl group donor. One noteworthy exception is seen in some bacteria, where the conserved tRNA methylation at m5U54 is added by the enzyme TrmFO using flavin adenine dinucleotide together with N5,N10-methylenetetrahydrofolate as the one-carbon donor. The minimalist bacterium Mycoplasma capricolum possesses two homologs of trmFO, but surprisingly lacks the m5U54 tRNA modification. We created single and dual deletions of the trmFO homologs using a novel synthetic biology approach. Subsequent analysis of the M. capricolum RNAs by mass spectrometry shows that the TrmFO homolog encoded by Mcap0476 specifically modifies m5U1939 in 23S rRNA, a conserved methylation catalyzed by AdoMet-dependent enzymes in all other characterized bacteria. The Mcap0476 methyltransferase (renamed RlmFO) represents the first folate-dependent flavoprotein seen to modify ribosomal RNA.

High-throughput comparison, functional annotation, and metabolic modeling of plant genomes using the PlantSEED resource.

The increasing number of sequenced plant genomes is placing new demands on the methods applied to analyze, annotate, and model these genomes. Today's annotation pipelines result in inconsistent gene assignments that complicate comparative analyses and prevent efficient construction of metabolic models. To overcome these problems, we have developed the PlantSEED, an integrated, metabolism-centric database to support subsystems-based annotation and metabolic model reconstruction for plant genomes. PlantSEED combines SEED subsystems technology, first developed for microbial genomes, with refined protein families and biochemical data to assign fully consistent functional annotations to orthologous genes, particularly those encoding primary metabolic pathways. Seamless integration with its parent, the prokaryotic SEED database, makes PlantSEED a unique environment for cross-kingdom comparative analysis of plant and bacterial genomes. The consistent annotations imposed by PlantSEED permit rapid reconstruction and modeling of primary metabolism for all plant genomes in the database. This feature opens the unique possibility of model-based assessment of the completeness and accuracy of gene annotation and thus allows computational identification of genes and pathways that are restricted to certain genomes or need better curation. We demonstrate the PlantSEED system by producing consistent annotations for 10 reference genomes. We also produce a functioning metabolic model for each genome, gapfilling to identify missing annotations and proposing gene candidates for missing annotations. Models are built around an extended biomass composition representing the most comprehensive published to date. To our knowledge, our models are the first to be published for seven of the genomes analyzed.

Integrative data-mining tools to link gene and function.

Information derived from genomic and post-genomic data can be efficiently used to link gene and function. Several web-based platforms have been developed to mine these types of data by integrating different tools. This method paper is designed to allow the user to navigate these platforms in order to make functional predictions. The main focus is on phylogenetic distribution and physical clustering tools, but other tools such as pathway reconstruction, gene fusions, and analysis of high-throughput experimental data are also surveyed.

Comparative genomics guided discovery of two missing archaeal enzyme families involved in the biosynthesis of the pterin moiety of tetrahydromethanopterin and tetrahydrofolate.

C-1 carriers are essential cofactors in all domains of life, and in Archaea, these can be derivatives of tetrahydromethanopterin (H(4)-MPT) or tetrahydrofolate (H(4)-folate). Their synthesis requires 6-hydroxymethyl-7,8-dihydropterin diphosphate (6-HMDP) as the precursor, but the nature of pathways that lead to its formation were unknown until the recent discovery of the GTP cyclohydrolase IB/MptA family that catalyzes the first step, the conversion of GTP to dihydroneopterin 2',3'-cyclic phosphate or 7,8-dihydroneopterin triphosphate [El Yacoubi, B.; et al. (2006) J. Biol. Chem., 281, 37586-37593 and Grochowski, L. L.; et al. (2007) Biochemistry46, 6658-6667]. Using a combination of comparative genomics analyses, heterologous complementation tests, and in vitro assays, we show that the archaeal protein families COG2098 and COG1634 specify two of the missing 6-HMDP synthesis enzymes. Members of the COG2098 family catalyze the formation of 6-hydroxymethyl-7,8-dihydropterin from 7,8-dihydroneopterin, while members of the COG1634 family catalyze the formation of 6-HMDP from 6-hydroxymethyl-7,8-dihydropterin. The discovery of these missing genes solves a long-standing mystery and provides novel examples of convergent evolutions where proteins of dissimilar architectures perform the same biochemical function.

Biosynthesis and function of posttranscriptional modifications of transfer RNAs.

Posttranscriptional modifications of transfer RNAs (tRNAs) are critical for all core aspects of tRNA function, such as folding, stability, and decoding. Most tRNA modifications were discovered in the 1970s; however, the near-complete description of the genes required to introduce the full set of modifications in both yeast and Escherichia coli is very recent. This led to a new appreciation of the key roles of tRNA modifications and tRNA modification enzymes as checkpoints for tRNA integrity and for integrating translation with other cellular functions such as transcription, primary metabolism, and stress resistance. A global survey of tRNA modification enzymes shows that the functional constraints that drive the presence of modifications are often conserved, but the solutions used to fulfill these constraints differ among different kingdoms, organisms, and species.

Biosynthesis of threonylcarbamoyl adenosine (t6A), a universal tRNA nucleoside.

The anticodon stem-loop (ASL) of transfer RNAs (tRNAs) drives decoding by interacting directly with the mRNA through codon/anticodon pairing. Chemically complex nucleoside modifications found in the ASL at positions 34 or 37 are known to be required for accurate decoding. Although over 100 distinct modifications have been structurally characterized in tRNAs, only a few are universally conserved, among them threonylcarbamoyl adenosine (t(6)A), found at position 37 in the anticodon loop of a subset of tRNA. Structural studies predict an important role for t(6)A in translational fidelity, and in vivo work supports this prediction. Although pioneering work in the 1970s identified the fundamental substrates for t(6)A biosynthesis, the enzymes responsible for its biosynthesis have remained an enigma. We report here the discovery that in bacteria four proteins (YgjD, YrdC, YjeE, and YeaZ) are both necessary and sufficient for t(6)A biosynthesis in vitro. Notably, YrdC and YgjD are members of universally conserved families that were ranked among the top 10 proteins of unknown function in need of functional characterization, while YeaZ and YjeE are specific to bacteria. This latter observation, coupled with the essentiality of all four proteins in bacteria, establishes this pathway as a compelling new target for antimicrobial development.

Functional promiscuity of the COG0720 family.

The biosynthesis of GTP derived metabolites such as tetrahydrofolate (THF), biopterin (BH(4)), and the modified tRNA nucleosides queuosine (Q) and archaeosine (G(+)) relies on several enzymes of the Tunnel-fold superfamily. A subset of these proteins includes the 6-pyruvoyltetrahydropterin (PTPS-II), PTPS-III, and PTPS-I homologues, all members of the COG0720 family that have been previously shown to transform 7,8-dihydroneopterin triphosphate (H(2)NTP) into different products. PTPS-II catalyzes the formation of 6-pyruvoyltetrahydropterin in the BH(4) pathway, PTPS-III catalyzes the formation of 6-hydroxylmethyl-7,8-dihydropterin in the THF pathway, and PTPS-I catalyzes the formation of 6-carboxy-5,6,7,8-tetrahydropterin in the Q pathway. Genes of these three enzyme families are often misannotated as they are difficult to differentiate by sequence similarity alone. Using a combination of physical clustering, signature motif, phylogenetic codistribution analyses, in vivo complementation studies, and in vitro enzymatic assays, a complete reannotation of the COG0720 family was performed in prokaryotes. Notably, this work identified and experimentally validated dual function PTPS-I/III enzymes involved in both THF and Q biosynthesis. Both in vivo and in vitro analyses showed that the PTPS-I family could tolerate a translation of the active site cysteine and was inherently promiscuous, catalyzing different reactions on the same substrate or the same reaction on different substrates. Finally, the analysis and experimental validation of several archaeal COG0720 members confirmed the role of PTPS-I in archaeosine biosynthesis and resulted in the identification of PTPS-III enzymes with variant signature sequences in Sulfolobus species. This study reveals an expanded versatility of the COG0720 family members and illustrates that for certain protein families extensive comparative genomic analysis beyond homology is required to correctly predict function.

Synergistic use of plant-prokaryote comparative genomics for functional annotations.

BACKGROUND: Identifying functions for all gene products in all sequenced organisms is a central challenge of the post-genomic era. However, at least 30-50% of the proteins encoded by any given genome are of unknown or vaguely known function, and a large number are wrongly annotated. Many of these 'unknown' proteins are common to prokaryotes and plants. We set out to predict and experimentally test the functions of such proteins. Our approach to functional prediction integrates comparative genomics based mainly on microbial genomes with functional genomic data from model microorganisms and post-genomic data from plants. This approach bridges the gap between automated homology-based annotations and the classical gene discovery efforts of experimentalists, and is more powerful than purely computational approaches to identifying gene-function associations. RESULTS: Among Arabidopsis genes, we focused on those (2,325 in total) that (i) are unique or belong to families with no more than three members, (ii) occur in prokaryotes, and (iii) have unknown or poorly known functions. Computer-assisted selection of promising targets for deeper analysis was based on homology-independent characteristics associated in the SEED database with the prokaryotic members of each family. In-depth comparative genomic analysis was performed for 360 top candidate families. From this pool, 78 families were connected to general areas of metabolism and, of these families, specific functional predictions were made for 41. Twenty-one predicted functions have been experimentally tested or are currently under investigation by our group in at least one prokaryotic organism (nine of them have been validated, four invalidated, and eight are in progress). Ten additional predictions have been independently validated by other groups. Discovering the function of very widespread but hitherto enigmatic proteins such as the YrdC or YgfZ families illustrates the power of our approach. CONCLUSIONS: Our approach correctly predicted functions for 19 uncharacterized protein families from plants and prokaryotes; none of these functions had previously been correctly predicted by computational methods. The resulting annotations could be propagated with confidence to over six thousand homologous proteins encoded in over 900 bacterial, archaeal, and eukaryotic genomes currently available in public databases.

Gcn4 misregulation reveals a direct role for the evolutionary conserved EKC/KEOPS in the t6A modification of tRNAs.

The EKC/KEOPS complex is universally conserved in Archaea and Eukarya and has been implicated in several cellular processes, including transcription, telomere homeostasis and genomic instability. However, the molecular function of the complex has remained elusive so far. We analyzed the transcriptome of EKC/KEOPS mutants and observed a specific profile that is highly enriched in targets of the Gcn4p transcriptional activator. GCN4 expression was found to be activated at the translational level in mutants via the defective recognition of the inhibitory upstream ORFs (uORFs) present in its leader. We show that EKC/KEOPS mutants are defective for the N6-threonylcarbamoyl adenosine modification at position 37 (t(6)A(37)) of tRNAs decoding ANN codons, which affects initiation at the inhibitory uORFs and provokes Gcn4 de-repression. Structural modeling reveals similarities between Kae1 and bacterial enzymes involved in carbamoylation reactions analogous to t(6)A(37) formation, supporting a direct role for the EKC in tRNA modification. These findings are further supported by strong genetic interactions of EKC mutants with a translation initiation factor and with threonine biosynthesis genes. Overall, our data provide a novel twist to understanding the primary function of the EKC/KEOPS and its impact on several essential cellular functions like transcription and telomere homeostasis.

A role for the universal Kae1/Qri7/YgjD (COG0533) family in tRNA modification.

The YgjD/Kae1 family (COG0533) has been on the top-10 list of universally conserved proteins of unknown function for over 5 years. It has been linked to DNA maintenance in bacteria and mitochondria and transcription regulation and telomere homeostasis in eukaryotes, but its actual function has never been found. Based on a comparative genomic and structural analysis, we predicted this family was involved in the biosynthesis of N(6)-threonylcarbamoyl adenosine, a universal modification found at position 37 of tRNAs decoding ANN codons. This was confirmed as a yeast mutant lacking Kae1 is devoid of t(6)A. t(6)A(-) strains were also used to reveal that t(6)A has a critical role in initiation codon restriction to AUG and in restricting frameshifting at tandem ANN codons. We also showed that YaeZ, a YgjD paralog, is required for YgjD function in vivo in bacteria. This work lays the foundation for understanding the pleiotropic role of this universal protein family.

Discovery and characterization of an amidinotransferase involved in the modification of archaeal tRNA.

The presence of the 7-deazaguanosine derivative archaeosine (G(+)) at position 15 in tRNA is one of the diagnostic molecular characteristics of the Archaea. The biosynthesis of this modified nucleoside is especially complex, involving the initial production of 7-cyano-7-deazaguanine (preQ(0)), an advanced precursor that is produced in a tRNA-independent portion of the biosynthesis, followed by its insertion into the tRNA by the enzyme tRNA-guanine transglycosylase (arcTGT), which replaces the target guanine base yielding preQ(0)-tRNA. The enzymes responsible for the biosynthesis of preQ(0) were recently identified, but the enzyme(s) catalyzing the conversion of preQ(0)-tRNA to G(+)-tRNA have remained elusive. Using a comparative genomics approach, we identified a protein family implicated in the late stages of archaeosine biosynthesis. Notably, this family is a paralog of arcTGT and is generally annotated as TgtA2. Structure-based alignments comparing arcTGT and TgtA2 reveal that TgtA2 lacks key arcTGT catalytic residues and contains an additional module. We constructed a Haloferax volcanii DeltatgtA2 derivative and demonstrated that tRNA from this strain lacks G(+) and instead accumulates preQ(0). We also cloned the corresponding gene from Methanocaldococcus jannaschii (mj1022) and characterized the purified recombinant enzyme. Recombinant MjTgtA2 was shown to convert preQ(0)-tRNA to G(+)-tRNA using several nitrogen sources and to do so in an ATP-independent process. This is the only example of the conversion of a nitrile to a formamidine known in biology and represents a new class of amidinotransferase chemistry.

Towards a systems approach in the genetic analysis of archaea: Accelerating mutant construction and phenotypic analysis in Haloferax volcanii.

With the availability of a genome sequence and increasingly sophisticated genetic tools, Haloferax volcanii is becoming a model for both Archaea and halophiles. In order for H. volcanii to reach a status equivalent to Escherichia coli, Bacillus subtilis, or Saccharomyces cerevisiae, a gene knockout collection needs to be constructed in order to identify the archaeal essential gene set and enable systematic phenotype screens. A streamlined gene-deletion protocol adapted for potential automation was implemented and used to generate 22 H. volcanii deletion strains and identify several potentially essential genes. These gene deletion mutants, generated in this and previous studies, were then analyzed in a high-throughput fashion to measure growth rates in different media and temperature conditions. We conclude that these high-throughput methods are suitable for a rapid investigation of an H. volcanii mutant library and suggest that they should form the basis of a larger genome-wide experiment.

A Gateway platform for functional genomics in Haloferax volcanii: deletion of three tRNA modification genes.

In part due to the existence of simple methods for its cultivation and genetic manipulation, Haloferax volcanii is a major archaeal model organism. It is the only archaeon for which the whole set of post-transcriptionally modified tRNAs has been sequenced, allowing for an in silico prediction of all RNA modification genes present in the organism. One approach to check these predictions experimentally is via the construction of targeted gene deletion mutants. Toward this goal, an integrative "Gateway vector" that allows gene deletion in H. volcanii uracil auxotrophs was constructed. The vector was used to delete three predicted tRNA modification genes: HVO_2001 (encoding an archaeal transglycosyl tranferase or arcTGT), which is involved in archeosine biosynthesis; HVO_2348 (encoding a newly discovered GTP cyclohydrolase I), which catalyzes the first step common to archaeosine and folate biosynthesis; and HVO_2736 (encoding a member of the COG1444 family), which is involved in N(4)-acetylcytidine (ac(4)C) formation. Preliminary phenotypic analysis of the deletion mutants was conducted, and confirmed all three predictions.

The universal YrdC/Sua5 family is required for the formation of threonylcarbamoyladenosine in tRNA.

Threonylcarbamoyladenosine (t(6)A) is a universal modification found at position 37 of ANN decoding tRNAs, which imparts a unique structure to the anticodon loop enhancing its binding to ribosomes in vitro. Using a combination of bioinformatic, genetic, structural and biochemical approaches, the universal protein family YrdC/Sua5 (COG0009) was shown to be involved in the biosynthesis of this hypermodified base. Contradictory reports on the essentiality of both the yrdC wild-type gene of Escherichia coli and the SUA5 wild-type gene of Saccharomyces cerevisiae led us to reconstruct null alleles for both genes and prove that yrdC is essential in E. coli, whereas SUA5 is dispensable in yeast but results in severe growth phenotypes. Structural and biochemical analyses revealed that the E. coli YrdC protein binds ATP and preferentially binds RNA(Thr) lacking only the t(6)A modification. This work lays the foundation for elucidating the function of a protein family found in every sequenced genome to date and understanding the role of t(6)A in vivo.

Biosynthesis of 7-deazaguanosine-modified tRNA nucleosides: a new role for GTP cyclohydrolase I.

Queuosine (Q) and archaeosine (G(+)) are hypermodified ribonucleosides found in tRNA. Q is present in the anticodon region of tRNA(GUN) in Eukarya and Bacteria, while G(+) is found at position 15 in the D-loop of archaeal tRNA. Prokaryotes produce these 7-deazaguanosine derivatives de novo from GTP through the 7-cyano-7-deazaguanine (pre-Q(0)) intermediate, but mammals import the free base, queuine, obtained from the diet or the intestinal flora. By combining the results of comparative genomic analysis with those of genetic studies, we show that the first enzyme of the folate pathway, GTP cyclohydrolase I (GCYH-I), encoded in Escherichia coli by folE, is also the first enzyme of pre-Q(0) biosynthesis in both prokaryotic kingdoms. Indeed, tRNA extracted from an E. coli DeltafolE strain is devoid of Q and the deficiency is complemented by expressing GCYH-I-encoding genes from different bacterial or archaeal origins. In a similar fashion, tRNA extracted from a Haloferax volcanii strain carrying a deletion of the GCYH-I-encoding gene contains only traces of G(+). These results link the production of a tRNA-modified base to primary metabolism and further clarify the biosynthetic pathway for these complex modified nucleosides.

Comparative genomics of bacterial and plant folate synthesis and salvage: predictions and validations.

BACKGROUND: Folate 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. RESULTS: Based 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. CONCLUSION: For 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.