A link between gut community metabolism and pathogenesis: molecular hydrogen-stimulated glucarate catabolism aids Salmonella virulence

Abstract

Glucarate, an oxidized product of glucose, is a major serum organic acid in humans. Still, its role as a carbon source for a pathogen colonizing hosts has not been studied. We detected high-level expression of a potential glucarate permease encoding gene gudT when Salmonella enterica serovar Typhimurium are exposed to hydrogen gas (H(2)), a gaseous by-product of gut commensal metabolism. A gudT strain of Salmonella is deficient in glucarate-dependent growth, however, it can still use other monosaccharides, such as glucose or galactose. Complementation of the gudT mutant with a plasmid harbouring gudT restored glucarate-dependent growth to wild-type (WT) levels. The gudT mutant exhibits attenuated virulence: the mean time of death for mice inoculated with WT strain was 2 days earlier than for mice inoculated with the gudT strain. At 4 days postinoculation, liver and spleen homogenates from mice inoculated with a gudT strain contained significantly fewer viable Salmonella than homogenates from animals inoculated with the parent. The parent strain grew well H(2)-dependently in a minimal medium with amino acids and glucarate provided as the sole carbon sources, whereas the gudT strain achieved approximately 30% of the parent strain's yield. Glucarate-mediated growth of a mutant strain unable to produce H(2) was stimulated by H(2) addition, presumably owing to the positive transcriptional response to H(2). Gut microbiota-produced molecular hydrogen apparently signals Salmonella to catabolize an alternative carbon source available in the host. Our results link a gut microbiome-produced diffusible metabolite to augmenting bacterial pathogenesis.

Keywords: carbon transport; gut microbiome; in vivo pathogen growth; metabolism and virulence; microbial carbon utilization.

Figure 1.

Comparison of aerobic growth of S. Typhimurium WT strain JSG210 (represented as circles) with RLK6/ΔgudT (represented as triangles) in minimal medium with 0.4% d-glucarate. Dashed line indicates WT growth without d-glucarate (control). Data points are the mean from three replicate serum bottles for each strain/condition. The standard deviation was less than 5% of the mean in every case, so that JSG210 with glucarate is greater than the lower two lines for all points at 4 h and greater at p ≤ 0.05 by Student's t-test.

Figure 2.

Comparison of H₂-facilitated anaerobic growth of S. Typhimurium strains. Anaerobic WT/JSG210 (represented as circles), RLK7/ΔhycC (represented as diamonds) and RLK6/ΔgudT (represented as triangles) in minimal medium with 0.4% d-glucarate. Solid lines indicate growth with added H₂ (20% v/v) and dashed lines indicate growth without added H₂. The standard deviation was less than 4% of the mean in every case, so that the added H₂ condition is significantly greater than without H₂ for each individual strain for all points at 4 h and greater at p ≤ 0.05 by Student's t-test. Without H₂, there was also a significant difference in JSG210 and the RLK7 mutant at 6 and 8 h points (p ≤ 0.05 by Student's t-test), while with H₂, the WT was significantly greater than strain RLK6 only.

Figure 3.

Comparison of virulence of S. Typhimurium strains JSG210/WT (represented as circles) and RLK6/ΔgudT (represented as triangles) in mice. The results shown are for 16 mice infected with each strain. The second experiment with eight mice infected with each strain showed similar results. A Wilcoxon rank-sum statistical analysis of these data was performed, testing that the distribution of dataset A (RLK6) is significantly shifted to the right of dataset B (WT), or H1: A > B using 16 data points for each strain. This test showed significance between the two groups at p ≤ 0.05 for a two-tailed test.