Metabolite Damage and Damage Control in a Minimal Genome

Abstract

Analysis of the genes retained in the minimized Mycoplasma JCVI-Syn3A genome established that systems that repair or preempt metabolite damage are essential to life. Several genes known to have such functions were identified and experimentally validated, including 5-formyltetrahydrofolate cycloligase, coenzyme A (CoA) disulfide reductase, and certain hydrolases. Furthermore, we discovered that an enigmatic YqeK hydrolase domain fused to NadD has a novel proofreading function in NAD synthesis and could double as a MutT-like sanitizing enzyme for the nucleotide pool. Finally, we combined metabolomics and cheminformatics approaches to extend the core metabolic map of JCVI-Syn3A to include promiscuous enzymatic reactions and spontaneous side reactions. This extension revealed that several key metabolite damage control systems remain to be identified in JCVI-Syn3A, such as that for methylglyoxal. IMPORTANCE Metabolite damage and repair mechanisms are being increasingly recognized. We present here compelling genetic and biochemical evidence for the universal importance of these mechanisms by demonstrating that stripping a genome down to its barest essentials leaves metabolite damage control systems in place. Furthermore, our metabolomic and cheminformatic results point to the existence of a network of metabolite damage and damage control reactions that extends far beyond the corners of it that have been characterized so far. In sum, there can be little room left to doubt that metabolite damage and the systems that counter it are mainstream metabolic processes that cannot be separated from life itself.

Keywords: comparative genomics; hydrolase; metabolite repair; metabolomics; minimal genome.

FIG 1

5-FCL activity is encoded by JCVISYN3A_0400. (A) Enzymatic source and repair of 5-CHO-THF. (B) Growth phenotype of a wild-type (WT) E. coli BW25113, a ΔygfA mutant, and a ΔygfA mutant expressing the JCVISYN3A_0443 gene on M9 minimal medium (0.4% glucose) with 20 mM NH₄Cl (plate 1) or 50 mM glycine (plate 2) as sole nitrogen source. Plates were incubated for 3 days at 37°C.

FIG 2

Predicted and validated redox buffering systems in JCVI-Syn3. (A) Candidates for H₂O₂ detoxification systems of JCVI-Syn3; those experimentally validated are shown by solid arrows, only the numbers of the locus tags are given, P is for protein, and R is for small molecule. The predicted source of reductant is NADPH. (B) CoADR Michaelis-Menten saturation curve for the determination of the K_m and k_cat for CoAD consumption. (C) CoADR is specific toward oxidized CoA with no activity toward other tested disulfides.

FIG 3

Heatmap including 15 metabolites from JCVI-Syn3A mutant metabolomic analysis with highest VIP scores. Samples and genotypes are represented in columns. High-intensity measurements compared to average intensity are red/yellow, and low-intensity measurements are represented by green/blue.

FIG 4

Predicted hydrolase of unknown function is clustered or fused to NadD in many Firmicutes. (A) Predicted NADP⁺ synthesis pathway in JCVI-Syn3. (B) Physical clustering and fusions of nadD and ykeK homologs in several Gram-positive bacteria. The RefSeq identifiers for the yqeK genes used in descending order are CAE77070, NP_390441.1, by AAW38265.1, AAO82560.1, and AAV61221.1.

FIG 5

Biochemical analysis of the NadD and YqeK activities. (A) Relative reaction rates of Bacillus subtilis and JCVI-Syn3.0 NadD enzymes with NaMN and various nucleotides, calculated as percentage of the canonical reaction with ATP for each NadD enzyme. Enzymes were incubated with 2 mM NTP, 0.5 mM NaMN, 4 mM MgCl₂, and 5 U/mL yeast inorganic pyrophosphatase for 5 min at 37°C. H230A has the conserved H in the active site of the YqeK domain mutated to ablate the HD activity and cleavage of nucleotides. (B) Activity of the expressed JCVI-Syn3.0 YqeK domain with different substrates. YqeK (0.2 μg) was incubated with 0.5 or 0.05 mM concentrations of the substrates, 1 mg/mL bovine serum albumin, and 2.0 mM MgCl₂ for 20 min at 37°C. Black bars are data for 0.5 mM substrates, and white bars are data for 0.05 mM substrates. (C) Mutation ratio on LB-rifampicin for ΔmutT strain with empty vector (pBAD24), ΔmutT strain with E. coli mutT in trans, and ΔmutT strain with either the nadD-yqeK fusion gene JCVISYN3A_0380 or the nadD or yqeK domain alone. *** indicates a P value of <0.001 with experiments performed with four biological replicates and four technical replicates.

FIG 6

Number of predicted potential metabolites arising from promiscuous enzymatic reactions and spontaneous/damage chemistry operating on known compounds in JCVI-Syn3 metabolism. Total predicted metabolites are shown, as well as the number of metabolites matching observed peaks (blue line) or ModelSEED compounds (green line). The x axis indicates the number of reaction steps explored outward from the known JCVI-Syn3 metabolism, while the y axis shows the number of new metabolites predicted with each new reaction step.

FIG 7

Map of predicted extensions to the JCVI-Syn3 model to push flux through as many observed peaks as possible. Reactions and metabolites are color coded as shown in the inset. Model reactions with no flux are black, and those with flux are magenta. Predicted and active reactions that are in the database are green, those that are novel and spontaneous are red, and those that are novel enzymatic ones are blue. All active predicted spontaneous reactions and nearly all active model reactions are shown on the map; some ModelSEED and predicted enzymatic reactions are excluded. The color code for metabolites is as follows: those absent in the mass spectrometry analysis are white, observed metabolites that are also in the model are yellow, those in the database are ocher, and those that are novel in themselves or in the way they are produced are brown. Most enzymatic reactions are identified by their EC numbers. Some reactants’ names have been omitted since they do not give relevant information. Common abbreviations have been used for the name labels. The map has been divided by panels shown on the figure’s background. These panels are labeled according to the major pathway they display. The complete metabolic map in interactive format (Escher map) is given in Supplemental data S5 at figshare (https://doi.org/10.6084/m9.figshare.20020574).

FIG 8

Characterization of JCVISYN3A_0400. (A) Growth of wild-type, ΔyajL, and ΔyajL ΔhchA strains; ΔyajL ΔhchA strain with hchA in trans; and ΔyajL ΔhchA strain with JCVISYN3A_0400 in trans. pUC19 was used as an empty vector. Each strain was tested in 5 replicates. Plates were incubated 2 days at 37°C in LB with agitation in a Bioscreen C device. OD, optical density. (B) Methylglyoxalase (MG) activity of JCVISYN3A_0400 compared to human DJ-1 (DJ1) and E. coli YajL. Conversion of methylglyoxal to l-lactate was measured in a coupled assay with l-lactate oxidase and Amplex red. Data were measured in triplicate with error bars shown (sometimes smaller than the symbol) and fitted using the Michaelis-Menten model. JCVISYN3A_0400 is a weak methylglyoxalase.