Integrative genome-scale metabolic analysis of Vibrio vulnificus for drug targeting and discovery

Abstract

Although the genomes of many microbial pathogens have been studied to help identify effective drug targets and novel drugs, such efforts have not yet reached full fruition. In this study, we report a systems biological approach that efficiently utilizes genomic information for drug targeting and discovery, and apply this approach to the opportunistic pathogen Vibrio vulnificus CMCP6. First, we partially re-sequenced and fully re-annotated the V. vulnificus CMCP6 genome, and accordingly reconstructed its genome-scale metabolic network, VvuMBEL943. The validated network model was employed to systematically predict drug targets using the concept of metabolite essentiality, along with additional filtering criteria. Target genes encoding enzymes that interact with the five essential metabolites finally selected were experimentally validated. These five essential metabolites are critical to the survival of the cell, and hence were used to guide the cost-effective selection of chemical analogs, which were then screened for antimicrobial activity in a whole-cell assay. This approach is expected to help fill the existing gap between genomics and drug discovery.

Figure 1

Reconstruction of genome-scale metabolic network and its use for drug targeting and discovery. (A) Genome-scale metabolic network of V. vulnificus CMCP6, VvuMBEL943, was reconstructed using genome annotation data, information from databases, literature, and experiments. As a result, the metabolic network has a coherently organized set of biochemical reactions along with information of their genes and proteins. (B) Once validated, it was employed for drug-targeting simulations using constraints-based flux analysis for predicting essential metabolites. In this study, consuming reactions associated with essential metabolites were experimentally validated by knocking out corresponding genes. (C) Finally, structural analogs of essential metabolites were collected from the chemical compound library, and used for drug discovery through whole-cell screening.

Figure 2

Radar plots of genomic differences among six representative Vibrio species. Blue and red lines indicate chromosome 1 and 2 of each species, respectively, for (A–E), while green line in (F) is for plasmid. Each plot reveals, for each organism, (A) chromosome size (bp), (B) percentage of G+C, (C) number of coding sequences (CDS), (D) average size of CDS (bp), (E) percentage of base pairs involved in encoding proteins, and (F) plasmid sizes (bp) in VFI and VHA (others do not have plasmids). Abbreviations are: VCH, V. cholerae N16961; VFI, V. fischeri ES114; VHA, V. harveyi ATCC BAA-1116; VPA, V. parahaemolyticus RIMD 2210633; VSP, V. splendidus LGP32s; VVU, V. vulnificus CMCP6. Detailed numbers for each category are also shown in Supplementary Table I.

Figure 3

Prediction of essential metabolites and systems-oriented drug targeting. On the top left, ovals and boxes indicate reactions and metabolites, respectively. Disrupted reactions are indicated with dotted lines along with crosses. Numbers next to each arrow indicate the number of essential metabolites survived after each filtering step, while list of metabolite abbreviations on the right side of each step shows the essential metabolites removed. Essential metabolites go through a series of filtering criteria. First, currency metabolites, which refer to metabolites that participate in many reactions, such as ATP and NADH, are removed. Then, essential metabolites consumed by multiple reactions are further selected in order to minimize the likelihood of resistance development while inhibiting the growth of the pathogen at sufficiently low concentration by simultaneously disturbing multiple targets. Finally, essential metabolites that also exist in human or that are associated with enzymes homologous to human proteins are removed in order to avoid any possible side effects in humans.

Figure 4

Initial set of 193 essential metabolites grouped into their relevant submetabolisms charted on the graph as functions of the portion of essential metabolites (colored bubbles), average number of surrounding reactions (y axis) and average essentiality of whole reactions constituting the submetabolism (x axis). Data for each submetabolism are labeled next to, or inside the bubble; the three numbers in sequence represent the average essentiality of metabolic reactions, the average number of surrounding reactions, and the bubble size (representing the portion of essential metabolites). Numbers in the parentheses indicate the number of essential metabolites in each submetabolism. The submetabolism (bubble) with the average essentiality of reactions closer to the value of 1 means that it has a higher portion of essential reactions. In this analysis, metabolism of cofactors and vitamins contained the largest number of essential metabolites, followed by amino acid and lipid metabolisms. Meanwhile, lipid metabolism topped average essentiality of reactions, while nucleotide metabolism had most abundant reactions surrounding essential metabolites.

Figure 5

Metabolic map showing genes knocked out for the validation of final drug targets, including VV10567, VV10580, VV11175, VV11568, VV11644, VV11691, and VV11866. All disrupted genes, displayed as colored lines on the metabolic map, were essential. Red and green lines indicate no mutant constructed and insertion mutant created with low frequency, respectively. Five essential metabolites are represented with their structures. Multiple arrows indicate reactions not shown in the metabolic pathway. Metabolite abbreviations can be found in Supplementary Table IV.

Figure 6

A hit compound 24837 screened from the whole-cell screening and its time-kill assay. (A) Structures of essential metabolite 4-aminobenzoate (PABA) and its structural analog 24837 are shown along with minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) values of the analog 24837 against V. vulnificus CMCP6. (B) Time-kill assays of compound 24837 and sulfamethoxazole, an existing chemotherapeutic agent that also interferes with folate biosynthetic pathway. It should be noted that the curve for 1 μg/ml of 24837 (yellow triangle) is covered with that for 0.5 μg/ml (closed rectangle) as they showed similar values. Data points and error bars indicate mean values±s.d.