The Thiamin-Requiring 3 Mutation of Arabidopsis 5-Deoxyxylulose-Phosphate Synthase 1 Highlights How the Thiamin Economy Impacts the Methylerythritol 4-Phosphate Pathway

The thiamin-requiring mutants of Arabidopsis have a storied history as a foundational model for biochemical genetics in plants and have illuminated the central role of thiamin in metabolism. Recent integrative genetic and biochemical analyses of thiamin biosynthesis and utilization imply that leaf metabolism normally operates close to thiamin-limiting conditions. Thus, the mechanisms that allocate thiamin-diphosphate (ThDP) cofactor among the diverse thiamin-dependent enzymes localized in plastids, mitochondria, peroxisomes, and the cytosol comprise an intricate thiamin economy. Here, we show that the classical thiamin-requiring 3 (th3) mutant is a point mutation in plastid localized 5-deoxyxylulose synthase 1 (DXS1), a key regulated enzyme in the methylerythritol 4-phosphate (MEP) isoprene biosynthesis pathway. Substitution of a lysine for a highly conserved glutamate residue (E323) located at the subunit interface of the homodimeric enzyme conditions a hypomorphic phenotype that can be rescued by supplying low concentrations of thiamin in the medium. Analysis of leaf thiamin vitamers showed that supplementing the medium with thiamin increased total ThDP content in both wild type and th3 mutant plants, supporting a hypothesis that the mutant DXS1 enzyme has a reduced affinity for the ThDP cofactor. An unexpected upregulation of a suite of biotic-stress-response genes associated with accumulation of downstream MEP intermediate MEcPP suggests that th3 causes mis-regulation of DXS1 activity in thiamin-supplemented plants. Overall, these results highlight that the central role of ThDP availability in regulation of DXS1 activity and flux through the MEP pathway.

Metabolic engineering provides insight into the regulation of thiamin biosynthesis in plants

Thiamin (or thiamine) is a water-soluble B-vitamin (B1), which is required, in the form of thiamin pyrophosphate (TPP), as an essential cofactor in crucial carbon metabolism reactions in all forms of life. To ensure adequate metabolic functioning, humans rely on a sufficient dietary supply of thiamin. Increasing thiamin levels in plants via metabolic engineering is a powerful strategy to alleviate vitamin B1 malnutrition and thus improve global human health. These engineering strategies rely on comprehensive knowledge of plant thiamin metabolism and its regulation. Here, multiple metabolic engineering strategies were examined in the model plant Arabidopsis thaliana. This was achieved by constitutive overexpression of the three biosynthesis genes responsible for B1 synthesis, HMP-P synthase (THIC), HET-P synthase (THI1) and HMP-P kinase/TMP pyrophosphorylase (TH1), either separate or in combination. By monitoring the levels of thiamin, its phosphorylated entities, and its biosynthetic intermediates, we gained insight into the effect of either strategy on thiamin biosynthesis. Moreover, expression analysis of thiamin biosynthesis genes showed the plant’s intriguing ability to respond to alterations in the pathway. Overall, we revealed the necessity to balance the pyrimidine and thiazole branches of thiamin biosynthesis and assessed its biosynthetic intermediates. Furthermore, the accumulation of non-phosphorylated intermediates demonstrated the inefficiency of endogenous thiamin salvage mechanisms. These results serve as guidelines in the development of novel thiamin metabolic engineering strategies.

The PLUTO plastidial nucleobase transporter also transports the thiamin precursor hydroxymethylpyrimidine

In plants, the hydroxymethylpyrimidine (HMP) and thiazole precursors of thiamin are synthesized and coupled together to form thiamin in plastids. Mutants unable to form HMP can be rescued by exogenous HMP, implying the presence of HMP transporters in the plasma membrane and plastids. Analysis of bacterial genomes revealed a transporter gene that is chromosomally clustered with thiamin biosynthesis and salvage genes. Its closest Arabidopsis homolog, the plastidic nucleobase transporter (PLUTO), is co-expressed with several thiamin biosynthetic enzymes. Heterologous expression of PLUTO in Escherichia coli or Saccharomyces cerevisiae increased sensitivity to a toxic HMP analog, and disrupting PLUTO in an HMP-requiring Arabidopsis line reduced root growth at low HMP concentrations. These data implicate PLUTO in plastidial transport and salvage of HMP.

A strictly monofunctional bacterial hydroxymethylpyrimidine phosphate kinase precludes damaging errors in thiamin biosynthesis

The canonical kinase (ThiD) that converts the thiamin biosynthesis intermediate hydroxymethylpyrimidine (HMP) monophosphate into the diphosphate can also very efficiently convert free HMP into the monophosphate in prokaryotes, plants, and fungi. This HMP kinase activity enables salvage of HMP, but it is not substrate-specific and so allows toxic HMP analogs and damage products to infiltrate the thiamin biosynthesis pathway. Comparative analysis of bacterial genomes uncovered a gene, thiD2, that is often fused to the thiamin synthesis gene thiE and could potentially encode a replacement for ThiD. Standalone ThiD2 proteins and ThiD2 fusion domains are small (∼130 residues) and do not belong to any previously known protein family. Genetic and biochemical analyses showed that representative standalone and fused ThiD2 proteins catalyze phosphorylation of HMP monophosphate, but not of HMP or its toxic analogs and damage products such as bacimethrin and 5-(hydroxymethyl)-2-methylpyrimidin-4-ol. As strictly monofunctional HMP monophosphate kinases, ThiD2 proteins eliminate a potentially fatal vulnerability of canonical ThiD, at the cost of the ability to reclaim HMP formed by thiamin turnover.

A cross-kingdom Nudix enzyme that pre-empts damage in thiamin metabolism

Genes specifying the thiamin monophosphate phosphatase and adenylated thiazole diphosphatase steps in fungal and plant thiamin biosynthesis remain unknown, as do genes for ThDP (thiamin diphosphate) hydrolysis in thiamin metabolism. A distinctive Nudix domain fused to Tnr3 (thiamin diphosphokinase) in Schizosaccharomyces pombe was evaluated as a candidate for these functions. Comparative genomic analysis predicted a role in thiamin metabolism, not biosynthesis, because free-standing homologues of this Nudix domain occur not only in fungi and plants, but also in proteobacteria (whose thiamin biosynthesis pathway has no adenylated thiazole or thiamin monophosphate hydrolysis steps) and animals (which do not make thiamin). Supporting this prediction, recombinant Tnr3 and its Saccharomyces cerevisiae, Arabidopsis and maize Nudix homologues lacked thiamin monophosphate phosphatase activity, but were active against ThDP, and up to 60-fold more active against diphosphates of the toxic thiamin degradation products oxy- and oxo-thiamin. Deleting the S. cerevisiae Nudix gene (YJR142W) lowered oxythiamin resistance, overexpressing it raised resistance, and expressing its plant or bacterial counterparts restored resistance to the YJR142W deletant. By converting the diphosphates of damaged forms of thiamin into monophosphates, the Tnr3 Nudix domain and its homologues can pre-empt the misincorporation of damaged diphosphates into ThDP-dependent enzymes, and the resulting toxicity.