Amino acid competition shapes Acinetobacter baumannii gut carriage

Abstract

Asymptomatic colonization is often critical for persistence of antimicrobial-resistant pathogens, such as Acinetobacter baumannii, and can increase the risk of clinical infections. However, the ecological factors shaping A. baumannii gut colonization remain unclear. We show that A. baumannii and other pathogenic Acinetobacter evolved to utilize the amino acid ornithine, a non-preferred carbon source, to compete with resident microbiota and persist in the gut in mice. A. baumannii encodes ornithine succinyltransferase (AstO) necessary for catabolizing ornithine, especially in conditions of increased microbial diversity. Supplemental dietary ornithine promotes long-term fecal shedding of A. baumannii. By contrast, supplementation of preferred carbon sources-monosodium glutamate or histidine-abolishes the requirement for ornithine catabolism. Additionally, A. baumannii gut carriage is higher in formula-fed human infants, who generally consume higher levels of protein, revealing dietary impacts on Acinetobacter colonization. Together, these results reveal that ornithine catabolism facilitates A. baumannii colonization, providing a reservoir for pathogen spread.

Keywords: Acinetobacter baumannii; antimicrobial resistance; carbon preference; dietary supplements; gut colonization; microbiome; microbiota; ornithine.

Figure 1.. A partial second AST pathway operon confers ornithine catabolism in pathogenic Acinetobacter species and contributes to A. baumannii gut colonization

(A) ast loci in the A. baumannii complex. (B) Schematic of the predicted arginine succinyltransferase pathway in A. baumannii. (C) A. baumannii WT, ΔastO, astA^{L125A H229A} and astA^{L125A H229A} ΔastO strains were grown in M9 minimal media with the indicated sole carbon source. Growth was monitored by optical density at 600 nm (OD₆₀₀) (n = 3, mean ± SD, experiments were repeated at least six times with similar results). (D) Phylogenetic species tree with inferred copy numbers of AstA/O and AstC/N (tree scale in amino acid substitutions). Max OD₆₀₀ of cultures grown with arginine or ornithine as the sole carbon or nitrogen source in M9 medium with succinate (glucose for E. coli) or NH₄ as controls. Growth was over 20 h (n = 3, mean ± SD, experiments were performed at least twice with similar results). (E) Prevalence of the ast2 operon in genomes of A. baumannii isolates. (F) Experimental design of post-abx A. baumannii gut colonization model. (G–H) A. baumannii CFU from fecal samples and in the cecum and colon (n = 10 mice combined from 2 independent experiments; p by Wilcoxon test with Holm-Sidak’s multiple comparisons). Lines connect CFU of strains enumerated from the same mouse. AST, arginine succinyltransferase; Orn, ornithine; α-KG, α-ketoglutarate; OD₆₀₀, optical density at 600 nm; CFU, colony-forming units; DPI, days post inoculation; LOD, approximate limit of detection. See also Figures S1 and S2 and Table S1.

Figure 2.. Supplemental dietary ornithine promotes long-term A. baumannii gut colonization and fecal shedding in mice

(A) Experimental design for ornithine supplementation and A. baumannii 17978 inoculation. (B) A. baumannii 17978 CFU from fecal samples. (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons; experiments were repeated twice with similar results.) (C) Experimental design for ornithine supplementation and A. baumannii AB5075 inoculation. (D) A. baumannii AB5075 CFU from fecal samples (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons). Lines connect CFU of both strains enumerated from the same mouse. CFU, colony-forming units; DPI, days post inoculation; LOD, approximate limit of detection. See also Figure S3.

Figure 3.. Microbiota diversity associates with A. baumannii AstO-dependent ornithine utilization and the gut amino acid metabolome

(A) α-Diversity was calculated from 16S rRNA gene sequencing at 0 and 9 DPI of A. baumannii WT and ΔastO (n = 10; data combined from mice in Figures 1G and 1H; mean ± SD are shown; p by one-way ANOVA with Sidak’s multiple comparisons). (B) Representative image of A. baumannii WT colonization at 10 DPI in the colon, visualized by MiPACT-HCR. Scale bar is 100 μm. Red, pan-bacteria HCR; blue, DAPI; magenta, WGA-lectin; green, anti-Acinetobacter HCR. From mice in Figures S4D and S4E. (C) Experimental design with germ-free mice. (D) A. baumannii CFU from fecal samples of germ-free mice (n = 6 female and n = 4 male mice, p by two-way ANOVA with Sidak’s multiple comparisons; experiment was repeated three times with similar results). (E) L-ornithine was quantified in chow and fecal samples (chow samples are n = 3 [LOD = 1 nmol/g], fecal samples are n = 10 mice shown in Figures 1G, 1H, and 3D [LOD = 5 nmol/g], mean ± SD, p by two-way ANOVA with Fisher’s least significant difference multiple comparisons only on fecal samples). (F) Principal-component analysis of amino acids in fecal samples at 0 and 10 DPI with chow samples shown in each plot (n = 10 mice shown in Figures 1G, 1H, and 3D, chow n = 3). Lines connect CFU of both strains enumerated from the same mouse. DPI, days post inoculation; MiPACT-HCR, microbial identification after passive CLARITY technique via hybridization chain reaction; WGA, wheat germ agglutinin; CFU, colony-forming units; LOD, approximate limit of detection; PC, principal component. See also Figure S4 and Table S2.

Figure 4.. Glutamate is a preferred carbon source for A. baumannii that promotes gut colonization by the ΔastO mutant

(A–C) A. baumannii strains were grown in M9 media with the indicated carbon sources. Growth was monitored by OD₆₀₀ (n = 3; mean ± SD, experiments were repeated at least four times with similar results). (D) Experimental design for monosodium glutamate supplementation. (E and F) CFU from fecal samples and the cecum and colon (n = 8–9 combined from 2 independent experiments, p by multiple Wilcoxon tests with Holm-Sidak’s multiple comparisons). (G) Total A. baumannii CFU were enumerated from fecal samples (n = 8–9, p by two-way ANOVA with Sidak’s multiple comparisons). (H) Experimental design for untreated mice supplemented with ornithine or glutamate. (I–K) A. baumannii CFU from fecal samples. (n = 4; p by two-way ANOVA with Sidak’s multiple comparisons.) (L) Total A. baumannii CFU from fecal samples (n = 4; p by one-way ANOVA with Tukey’s multiple comparisons). Lines connect CFU of both strains enumerated from the same mouse. CFU, colony-forming units; DPI, days post inoculation; LOD, approximate limit of detection. See also Figure S5.

Figure 5.. Histidine rescue of A. baumannii ΔastO gut colonization requires histidine catabolism

(A) A. baumannii was grown in M9 media with the indicated carbon sources. Growth was monitored by OD₆₀₀ (n = 3; mean ± SD, experiment was repeated 4 times with similar results). (B) A. baumannii strains were grown in M9 media with the indicated sole carbon source. Growth was monitored by OD₆₀₀ (n = 3; mean ± SD, experiments were repeated at least twice with similar results). (C) Experimental design for hutH mutant gut colonization with histidine supplementation. (D) Post-abx mice were co-inoculated with A. baumannii WT and ΔastO. CFU from fecal samples at 1, 4, and 9 DPI (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons). (E) WT A. baumannii CFU from fecal samples with and without histidine supplementation (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons). (F) Post-abx mice were co-inoculated with A. baumannii ΔhutH and ΔhutHΔastO. CFU from fecal samples (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons). (G) ΔhutH single mutant A. baumannii CFU from fecal samples with and without histidine supplementation (n = 5, p by two-way ANOVA with Sidak’s multiple comparisons). Lines connect CFU of both strains enumerated from the same mouse. OD₆₀₀, optical density at 600 nm; Orn, ornithine; DPI, days post inoculation; LOD, approximate limit of detection. See also Figure S6.

Figure 6.. A. baumannii abundance in fecal samples from healthy human 4-month-old infants is increased with formula feeding

(A–C) Prevalence of A. baumannii, relative abundance of A. baumannii and microbiota α-diversity in the gut of infants age 1, 4, 12, and 24 months by shotgun metagenomic sequencing (1 month, n = 104; 4 months, n = 98; 12 months, n = 104; 24 months n = 88; p by Kruskal-Wallis with Dunn’s post hoc test). (D–F) Prevalence of A. baumannii, relative abundance of A. baumannii and microbiota α-diversity in the gut of infants across feeding types at 1 and 4 months (1 month formula, n = 45; 1 month breastmilk, n = 30; 1 month mixed, n = 29; 4 months formula, n = 62; 4 months breastmilk, n = 17; 4 months mixed, n = 19; p by Kruskal-Wallis with Dunn’s post hoc test). Boxplots: center line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; points, outliers.