NSF Grant #0114117

Principal Investigators:
Andrew D. Hanson            University of Florida      adha@mail.ifas.ufl.edu
Yair Shachar-Hill               University of Michigan   yairhill@msu.edu
 

The network of one-carbon (C1) reactions provides C1 units for use in many biosynthetic reactions. It is crucial to plant metabolism, but many of the network's enzymes are known poorly or not at all. In this project, functions are being determined for genes that putatively encode: 10-formyl-tetrahydrofolate (THF) deformylase, sarcosine oxidase, formamidase, 5-formyl-THF cycloligase, and methionine gamma-lyase (Fig.1). For all these genes "determining function" means finding the biochemical activities and subcellular location of the encoded proteins. This has been done for the 5-formyl-THF cycloligase (Roje S. et al., 2002). In addition, isolation of Arabidopsis knockout mutants or RNAi transgenic plants (Goyer A. et al., 2004) are used to determine the effects of the mutations on C1 metabolism using NMR, mass spectral, and biochemical procedures. GenBank numbers for the network of C1 genes are in Table I. The broader impact is three-fold. First, on crop improvement. Many efforts to genetically engineer plants for human benefit involve changes to C1 metabolism, making it vital to understand C1 metabolism so as to engineer it successfully. Second, on basic plant biochemistry, because C1 metabolism is perhaps the least well understood area of plant primary metabolism despite its central position in processes such as photorespiration, lignification, and alkaloid synthesis. Third, on biochemistry in general, because the plant C1 metabolic network is special, not merely a minor variation on those in yeast, liver, or bacteria.
 
 

  The plant C1 metabolism network as deduced from biochemical and genomic data.  Main sources of C1 units are boxed. Enzymes are numbered. Met, methionine; Hcy, homocysteine.  Enzymes: 1, formate dehydrogenase; 2, S-formylglutathione hydrolase; 3, formaldehyde dehydrogenase; 4, sarcosine oxidase; 5, 10-formyl-THF deformylase; 6, 10-formyl-THF synthetase; 7, Met-tRNA transformylase; 8, polypeptide deformylase; 9, formamidase; 10, GAR transformylase; 11, AICAR transformylase; 12 & 13, bifunctional 5,10-methenyl-THF cyclohydrolase/5,10-methylene-THF dehydrogenase; 14, 5-formyl-THF cycloligase; 15, 5-formimino-THF cyclodeaminase; 16, glutamate formiminotransferase; 17, glycine formiminotransferase; 18, SHMT; 19, GDC; 20, 5,10-methylene-THF reductase; 21, thymidylate synthase/dihydrofolate reductase; 22, ketopantanoate hydroxymethyltransferase; 23, Met synthase; 24, Met gamma-lyase; 25, AdoMet synthetase; 26, AdoHcy hydrolase; 27, Hcy methyltransferase; 28, Met methyltransferase. Enzymes indicated in bold are the subject of this proposal.
 
 

(ND): Non Determined
(D): Determined
* Predicted using TargetP, Predotar, and PSORT; the predictions of all three programs agreed in most cases.
 
 

1. 5-Formyl-THF cycloligase

5-Formyl-THF cycloligase (5-FCL) catalyzes the conversion of 5-formyl-THF (5-CHO-H4PteGlun) to 5,10-methenyl-THF and is considered to be the main means whereby 5-CHO-H4PteGlun is metabolized in mammals, yeast, and bacteria. 5-CHO-H4PteGlun is known to occur in plants and to be highly abundant in leaf mitochondria. Genomics-based approaches identified Arabidopsis and tomato cDNAs encoding proteins homologous to 5-FCLs of other organisms but containing N-terminal extensions with the features of mitochondrial targeting peptides. These homologs were shown to have 5-FCL activity by characterizing recombinant enzymes produced in Escherichia coli and by functional complementation of a yeast fau1 mutation with the Arabidopsis 5-FCL cDNA. The recombinant Arabidopsis enzyme is active as a monomer, prefers the penta- to the monoglutamyl form of 5-CHO-H4PteGlun, and has kinetic properties broadly similar to those of 5-FCLs from other organisms. Enzyme assays and immunoblot analyses indicated that 5-FCL is located predominantly if not exclusively in plant mitochondria and that the mature, active enzyme lacks the putative targeting sequence. Serine hydroxymethyltransferase (SHMT) from plant mitochondria was shown to be inhibited by 5-CHO-H4PteGlun, as are SHMTs from other organisms. Since mitochondrial SHMT is crucial to photorespiration, 5-FCL may help prevent 5-CHO-H4PteGlun from reaching levels that would inhibit this process. Consistent with this possibility, 5-FCL activity was far higher in leaf mitochondria than root mitochondria.

To better understand the function of 5-FCL in plants, a T-DNA knockout mutant for 5-FCL of Arabidopsis (Torrey Mesa Research Institute, line 28D07) was characterized. This mutant had a 5-CHO-H4PteGlun content of 1.04 nmol g-1 fresh weight under normal growing conditions, which was three-fold higher than wild-type plants (0.35 nmol g-1 fresh weight) but the total folate content was the same in knockout and wild-type. Glycine content was ~19-fold higher in the mutant compared to wild-type plants under normal conditions, and ~46-fold higher under photorespiratory conditions, but only two-fold higher under non-photorespiratory conditions. Serine content varied less between wild-type and mutant, with a maximum increase of 1.7-fold under photorespiratory conditions. When plants were fed with 5-CHO-THF under normal conditions, glycine accumulated ~6-fold in the mutant, and ~16-fold in the wild-type, confirming the impact of 5-CHO-THF content on glycine to serine fluxes. These results show that 5-CHO-H4PteGlun has an important role in the regulation of glycine to serine fluxes through serine hydroxymethyltransferase in photorespiration.
 

2. Sarcosine Oxidase
Sarcosine oxidase (SOX) is known as a peroxisomal enzyme in mammals, and as a sarcosine-inducible enzyme in soil bacteria. Its presence in plants was unsuspected until the Arabidopsis genome was found to encode a protein (AtSOX) with ~33% sequence identity to mammalian and bacterial SOXs. When overexpressed in Escherichia coli, AtSOX enhanced growth on sarcosine as sole nitrogen source, showing that it has SOX activity in vivo, and the recombinant protein catalyzed the oxidation of sarcosine to glycine, formaldehyde, and H2O2 in vitro. AtSOX also attacked other N-methyl amino acids and, like mammalian SOXs, catalyzed the oxidation of L-pipecolate to Delta1-piperideine-6-carboxylate. Like bacterial monomeric SOXs, AtSOX was active as a monomer, contained FAD covalently bound to a cysteine residue near the C-terminus, and was not stimulated by tetrahydrofolate. Although AtSOX lacks a typical peroxisome-targeting signal, in-vitro assays established that it is imported into peroxisomes. Quantitation of mRNA showed that AtSOX is expressed at a low level throughout the plant, and is not sarcosine-inducible. Consistent with a low level of AtSOX expression, Arabidopsis plantlets slowly metabolized supplied [14C]sarcosine to glycine and serine. GC-MS analysis revealed low levels of pipecolate but not sarcosine in wild type Arabidopsis, and showed that pipecolate but not sarcosine accumulated six-fold when AtSOX expression was suppressed by RNAi. Moreover, the pipecolate catabolite alpha-aminoadipate decreased 30-fold in RNAi plants. These data indicate that pipecolate is the endogenous substrate for SOX in plants, and suggest that plants utilize exogenous sarcosine opportunistically, sarcosine being a common soil metabolite.
 

3. 10-formyl-THF-deformylase
Two genes encoding putative 10-formyl-THF-deformylases (purU) were identified in the genome of Arabidopsis. Attempts to produce an enzymatically active recombinant protein in E. coli and yeast have been unsuccessful. As an alternative to better understand the role of these proteins, we looked for T-DNA knockouts mutants of Arabidopsis in different collections. One KO line for each of the purU genes (At5g47435 gene, line 169F06 from the Torrey Mesa Research Institute collection; At4g17360 gene, line SALK_062946 from the Salk Institute collection) were found. Both mutants had a single T-DNA insertion as determined by southern blot analysis and did not express mRNA of either purU gene as determined by Real-Time Quantitative RT-PCR. None of these mutants showed a visible phenotype.

The mutant lines were crossed and the progeny were self-pollinated. Homozygous lines for both mutations were identified by PCR screening. This double mutant grows normally under standard conditions, but produces only non-viable shrunken seeds.
 

4. Methionine gamma-lyase
Methionine gamma-lyase (MGL) catalyzes the conversion of L-methionine to alpha-ketobutyrate, methanethiol, and ammonia. We found in the Arabidopsis genome a gene (At1g64660) encoding a protein (AtMGL) with ~35% identity to bacterial and protozoon MGLs. When overexpressed in Escherichia coli, AtMGL allowed growth on L-methionine as sole nitrogen source. GC analysis showed that this overexpressing E. coli but not the control emits methanethiol. These results showed that AtMGL has MGL activity in vivo leading to the production of ammonia and methanethiol. The recombinant protein was purified and exhibited a typical spectrum of 5’-pyridoxal phosphate (PLP)-dependent proteins. The recombinant protein catalyzes alpha,gamma- and ß,gamma-elimination reactions of a number of derivatives of methionine and cysteine. Quantitation of mRNA showed that the AtMGL gene is expressed in all organs tested, and that AtMGL mRNA levels in leaves did not change after spraying Arabidopsis plantlets with L-methionine or homocysteine.

We are currently investigating the methanethiol and dimethyldisulfide emission from both wild type and knockout (Line SALK_040380 from the Salk Institute) plants.
 

5. Formamidase
Formamidase catalyzes the conversion of formamide to formate and ammonia. Two genes encoding putative formamidase were identified in the Arabidopsis genome (At4g37550, FMD1; At4g37560; FMD2). When over-expressed in E. coli, both FMDs allowed growth on formamide as sole nitrogen source, but FMD2 with a better efficiency. However, only FMD1 allowed E. coli to grow on acetamide as sole nitrogen source. His-tagged recombinant proteins were produced in E. coli, and purified on a nickel column to homogeneity. Cross-linking experiments showed that both proteins exist as a dimer. Each subunit has a molecular mass of ~51 kDa. Both proteins had formamidase activity in vitro, but FMD2 activity was ~10 times higher than FMD1.

To determine the role of these FMDs in planta, RNAi transgenic plants were generated, and are now being investigated.
 
 

The checked materials (*) are available from our laboratory upon request (adha@mail.ifas.ufl.edu or agoyer@mail.ifas.ufl.edu)
 
 
 

Goyer A, Johnson TL, Olsen LJ, Collakova E, Shachar-Hill Y, Rhodes D, Hanson AD (2004)
Characterization and metabolic function of a peroxisomal sarcosine and pipecolate oxidase from Arabidopsis
J. Biol. Chem. 279: 16947-16953

Sanja Roje, Machhindra T. Janave, Michael J. Ziemak, and Andrew D. Hanson (2002)
Cloning and characterization of mitochondrial 5-formyltetrahydrofolate cycloligase from higher plants
J Biol Chem. 2002 Aug; 277(45):42748-42754