Ethanolamine is a valuable nutrient source that impacts Clostridium difficile pathogenesis

Abstract

Clostridium (Clostridioides) difficile is a gastrointestinal pathogen that colonizes the intestinal tract of mammals and can cause severe diarrheal disease. Although C. difficile growth is confined to the intestinal tract, our understanding of the specific metabolites and host factors that are important for the growth of the bacterium is limited. In other enteric pathogens, the membrane-derived metabolite, ethanolamine (EA), is utilized as a nutrient source and can function as a signal to initiate the production of virulence factors. In this study, we investigated the effects of ethanolamine and the role of the predicted ethanolamine gene cluster (CD1907-CD1925) on C. difficile growth. Using targeted mutagenesis, we disrupted genes within the eut cluster and assessed their roles in ethanolamine utilization, and the impact of eut disruption on the outcome of infection in a hamster model of disease. Our results indicate that the eut gene cluster is required for the growth of C. difficile on ethanolamine as a primary nutrient source. Further, the inability to utilize ethanolamine resulted in greater virulence and a shorter time to morbidity in the animal model. Overall, these data suggest that ethanolamine is an important nutrient source within the host and that, in contrast to other intestinal pathogens, the metabolism of ethanolamine by C. difficile can delay the onset of disease.

Figure 1. C. difficile metabolizes ethanolamine as a primary nutrient source

Growth curve of strain A) 630Δerm or B) R20291 in modified minimal medium (MMM) or MMM supplemented with 15 mM ethanolamine (EA) and/or 5 mM D-glucose. The average of three independent biological replicates and standard deviations are shown.

Figure 2. Expression of ethanolamine utilization genes in C. difficile

A) Ethanolamine utilization (eut) gene cluster in C. difficile (eutG-eutQ). The eut gene cluster contains 19 predicted open reading frames (predicted functions listed in File S7). Genes in blue encode predicted metabolic enzymes, green encode regulatory factors, yellow are predicted microcompartment structural proteins, purple encode the putative transporter and white (hyp) is of unknown function. The putative factor downstream of eutQ (R, green stripes) is a MarR-family transcriptional regulator encoded in most sequenced C. difficile isolates downstream of eutQ, but is not found in strain 630. Suspected promoters and terminators are shown as arrows and lollipops, respectively. (B) qRT-PCR analysis of putative eut gene expression in strain 630Δerm grown in 70:30 liquid medium with and without the addition of 15 mM ethanolamine (EA). Samples were harvested for RNA isolation two hours after the entry into stationary phase (T₂). The means and standard error of the means of three biological replicates are shown. *, p ≤ 0.05 by Student’s two-tailed t-test.

Figure 3. eut gene expression increases in response to ethanolamine, but expression is repressed during stationary phase if ethanolamine is absent

qRT-PCR analysis of eutA expression of 630Δerm grown in 70:30 liquid medium with and without the addition of 15 mM ethanolamine (EA). Expression of eutA in 630Δerm grown in medium with and without EA was compared at each timepoint using the Student’s two-tailed t-test. The means and standard error of the means of three biological replicates are shown. * indicates p ≤ 0.05.

Figure 4. The eutA operon, and not the eutG transcriptional unit, is essential for ethanolamine metabolism in C. difficile

Growth curve with 630Δerm (black squares), eutA (MC394; blue circles) and the eutG strain (MC346; orange triangles) grown in modified minimal medium (MMM; open symbols) and with the addition of 15 mM ethanolamine (EA; filled symbols). Averages and standard deviations of three independent experiments shown.

Figure 5. Insertional disruption of eutG does not affect transcription of downstream genes

qRT-PCR analysis of eutG, eutS, eutP, eutV and eutW expression in 630Δerm and the eutG mutant (MC346) grown in 70:30 broth with and without the addition of 15 mM ethanolamine (EA). Samples for RNA isolation were collected two hours after the transition into stationary phase (T₂). The means and standard error of the means for three biological replicates are shown. Gene expression of 630Δerm and the eutG mutant were compared for each condition using the Student’s two-tailed t-test. * indicates p ≤ 0.05.

Figure 6. A eutA mutant is more virulent in the hamster model of CDI

A) Kaplan-Meier survival curve representing the results from three independent experiments of Syrian golden hamsters infected with C. difficile strain 630Δerm (n=17) or the eutA mutant (MC394; n=17). The mean times to morbidity were 56 +/− 39.6 h for 630Δerm and 33.7 +/− 4.3 h for the eutA mutant (P ≤ 0.01, log rank test). B) Total CFU of C. difficile were enumerated from hamster cecal samples collected post-mortem. Solid lines represent the median CFU for each strain and the dotted line denotes the limit of detection (2 × 10¹ CFU ml⁻¹). CFU for the mutant and parent strains were compared using the Student’s two-tailed t test (no statistically significant differences were observed).

Figure 7. Ethanolamine metabolism does not impact expression of known virulence factors in vitro

qRT-PCR analysis of A) tcdA, B) tcdB, C) pilA1 and D) fliC expression in 630Δerm and the eutA mutant (MC394) grown in 70:30 broth with and without the addition of 15 mM EA. Samples for RNA isolation were collected during logarithmic growth (Log, OD₆₀₀ of 0.5) and four hours after the transition into stationary phase (T4, late stationary). The means and standard error of the means of four biological replicates are shown. Gene expression of 630Δerm and the eutA mutant were compared at each timepoint and condition using the Student’s two-tailed t-test. * indicates p ≤ 0.05 (no significant differences were observed).