Subsystem Info

Last Modified:6-20-2013:13:16:2
DescriptionDESCRIPTION (1/5/2008)

YgfZ proteins belong to COG0354; they are distinguished from similar proteins by the motif KGC[YF]xGQE about 100 residues from the C-terminus.

1. Prior evidence that YgfZ/COG0354 members are folate-related proteins

a. Primary sequence similarity to GcvT (glycine cleavage complex T protein, which binds folate).
b. Protein fold predictions for COG0354 from all taxa show high similarity to GcvT, dimethylglycine oxidase, and sarcosine oxidase structures (as well as to YgfZ of E. coli)
c. The E. coli YgfZ crystal structure (1) indicates a folate-binding domain. Intrinsic tryptophan fluorescence data appeared to confirm binding of folic acid or tetrahydrofolate (THF) - but the experiments were flawed and we are unable to repeat them (see Verification Notes, below).
d. Among archaea, YgfZ occurs only in Halobacteria (which almost the only ones to produce folates).

2. Reported characteristics of YgfZ proteins

a. Overall sequence identity among diverse COG0354 members is typically no more than 25%.
b. The only absolutely conserved region is an octapeptide motif KGC[YF]xGQE motif that lies in an interdomain linker on the surface of the protein, not in the folate-binding site (1). This loop was suggested to be an interaction motif, but a structural rearrangement could conceivably bring the loop to the folate-binding site (1).
c. The YgfZ protein is induced in E. coli by plumbagin exposure (plumbagin generates ROS and is only toxic to E. coli in presence of oxygen). Other plumbagin-induced proteins included superoxide dismutase, AhpC, and Nfo (all oxidation-stress related) (2).
d. The yeast YgfZ homolog (Iba57p, formerly Caf17p) is a mitochondrial protein that physically interacts in a complex with the ISC Fe/S assembly proteins Isa1p and Isa2p (3).

3. Reported characteristics of ygfZ mutants

a. The E. coli ygfZ disruptant grows slowly, especially at low temperature. The level of 2-methylthio N6-isopentenyladenosine (ms2i6A) is reduced in the disruptant, and its precursor N6-isopentenyladenosine (i6A) accumulates (4). MiaB is a radical SAM Fe/S enzyme that mediates the posttranscriptional methylthiolatation of i6A. TrmE (= mnmE) mutations suppressed the ygfZ phenotype (4). TrmE is a folate-dependent protein that normally transfers a one-carbon unit to modify uridine bases in tRNA.
b. Yeast iba57 mutants are petite, cannot grow on non-fermentable carbon sources, and lack activity of aconitase and radical SAM type Fe/S proteins (3).

5. Essentiality data. ygfZ genes are essential or important in all organisms tested:

E. coli – deleterious (see ref. 4 and Verification Notes)
Mycobacterium tuberculosis – essential
Haemophilus influenzae – essential
Pseudomonas aeruginosa - essential
Yeast – petite phenotype
Arabidopsis - mitochondrial ygfZ essential (see Verification Notes)

6. Clustering patterns & distribution of ygfZ genes

There are clear associations between YgfZ, Fe/S proteins, and aerobic conditions:

a. ygfZ genes cluster with diverse Fe/S enzymes and related proteins at a frequency about 10-fold higher than predicted for random associations. For specific examples see Methylococcus capsulatus, Rubrobacter xylanophilus, Stenotrophomonas maltophila, Buchnera aphidicola, Candidatus Pelagibacter ubique.
b. ygfZ genes co-occur with genes encoding scaffold proteins of the HesB/IscA/SufA family in 90% of 570 genomes examined (cf complex formation in yeast between Iba57p and ISC proteins).
c. Only 15% of anaerobes have YgfZ versus 63% of aerobes. Of the 37% of aerobes that lack YgfZ, 22% have few Fe/S proteins, so presumably have no need of YgfZ.

7. Plant ygfZ genes

a. Arabidopsis and other higher plants have two YgfZ proteins, one predicted to be mitochondrial, the other plastidial (At4g12130 and At1g60990, respectively). Phylogenetic analysis groups At4g12130 with alpha-Proteobacteria, and At1g60990 with Cyanobacteria (consistent with mitochondrial and plasidial locations).
b. Mitochondrial At4g12130 has an unusual expression pattern – much more strongly in pollen and late seed development than elsewhere (these are periods of desiccation and oxidative stress). At1g60990 is expressed at low levels in aerial organs (consistent with a plastid location).
c. At4g13130 expression is strongly positively correlated with Ferritin2 (an iron storage protein) and with Frataxin (a mitochondrial protein that mediates Fe delivery during Fe/S cluster biogenesis). At1g60990 expression is strongly correlated with chloroplast biogenesis proteins.

8. Occurrence of YgfZ in other Eukaryotes

Present in: Plants – mitochondrial and chloroplast copies
Yeast & other fungi – mitochondrial
Insects – mitochondrial
Vertebates - mitochondrial


1. Teplyakov A, Obmolova G, Sarikaya E, Pullalarevu S, Krajewski W, Galkin A, Howard AJ, Herzberg O, Gilliland GL (2004) Crystal structure of the YgfZ protein from Escherichia coli suggests a folate-dependent regulatory role in one-carbon metabolism. J Bacteriol 186: 7134-40

2. Chen JW, Sun CM, Sheng WL, Wang YC, Syu WJ (2006) Expression analysis of up-regulated genes responding to plumbagin in Escherichia coli. J Bacteriol 188: 456-63

3. Gelling C, Dawes IW, Richhardt N, Lill R, Mühlenhoff U (2008) Mitochondrial Iba57p is required for Fe/S cluster formation on aconitase and activation of radical SAM enzymes. Mol Cell Biol 28: 1851-61

4. Ote T, Hashimoto M, Ikeuchi Y, Su'etsugu M, Suzuki T, Katayama T, Kato J (2006) Involvement of the Escherichia coli folate-binding protein YgfZ in RNA modification and regulation of chromosomal replication initiation. Mol Microbiol 59: 265-75


1. YgfZ is an ancient, folate-dependent protein involved in the assembly or repair of Fe/S clusters, particularly under oxidative stress. YgfZ may function to strip formaldehyde adducts from the cysteine residues that ligand the clusters, and transfer them to THF. One or more residues in the conserved loop could initially accept the formaldehyde unit, then transfer it to THF.

In favor of this conjecture:
a. Several residues in the conserved loop (Lys, Cys, Gln) can form formaldehyde adducts singly or together. Of these, Lys is probably the most reactive.
b. TrmE, which exacerbates the effect of inactivating YgfZ, is most probably a formaldehyde donor.
c. YgfZ is closest in structure to proteins that accept formaldehyde units.
d. Many protein side chains (especially Cys and Lys) form reversible formaldehyde adducts, their reactivity being determined by sequence context.
e. Cys adduct formation would block Fe/S cluster binding.
f. Oxidative stress may augment formaldehyde production (e.g. by lipid peroxidation) or impair formaldehyde detoxification (by depleting glutathione) (See e.g. A one-carbon modification of protein lysine associated with elevated oxidative stress in human substantia nigra. Floor E et al. J Neurochem. 2006 97:504-14; Formaldehyde in Human Cancer Cells: Sensitive Detection by Distillation Sampling-Chemical Ionization Mass Spectrometry and Prognostic Implications. Kato S et al. Anal. Chem., 2001, 73, 2992-2997)

2. Another hypothesis is that YgfZ is a subunit of an oxidative demethylase (cf. sarcosine and dimethylglycine oxidases). In favor of this, COG0354 is immediately upstream of a DadA homolog in 3 Psychrobacter genomes. But homology with these COG0354 proteins drops off very fast, and they are N-terminally truncated relative to other COG0354 proteins, i.e. they are outliers.


1. E. coli YgfZ was reported to bind folic acid and THF based on tryptophan fluorescence quenching. However, no correction was made for the inner filter effect, or for loss of fluorescence signal due to denaturation attendant upon mixing the sample after the sequential addition of ligand. When these factors are taken into account, the apparent fluorescence quenching by folate disappears. Alexey Teplyakov agreed that the original data were defective, and wrote (4/4/2008) "Let's put it simple. Forget about our experiment. You easily convinced me that it wasn't done properly."

2. The Arabidopsis mitochondrial YgfZ (At4g12130) knockout is lethal. The chloroplastic YgfZ (At1g60990 knockout) has yet to be tested.

3. The E. coli folE knockout (which completely lacks folates) has impaired MiaB activity (as measured by the ms2i6A/i6A ratio), similar to the ygfZ knockout. This indicates that a folate is required for YgfZ action.

4. The E. coli knockout has reduced activities of several Fe/S enzymes including succinate dehydrogenase, fumarase A+B, and DMSO reductase, especially when exposed to oxidative stress.

5. The E. coli ygfZ knockout is more sensitive to oxidative stress (imposed with plumbagin, phenazine methosulfate, or methyl viologen) than the wild type. The ygfZ knockout also grows poorly on minimal media. These phenotypes can be complemented by ygfZ genes from Arabidopsis.

6. Replacing the Cys residue in the conserved loop of E. coli YgfZ with Ala reduces but does not abolish ability to complement the ygfZ deletion. This makes involvement of the Cys residue in persulfide formation unlikely, but is compatible with the hypothesis of formaldehyde binding (in which the Cys residue is the lesser of two prime candidates).

Classification:Cofactors, Vitamins, Prosthetic Groups, Pigments
Folate and pterines