The colonization of the human gut microbiome by antibiotic-resistant Enterobacteriaceae strains is of peculiar concern in clinical practice, as it exposes to the risk of spread within the community, increasing the subsequent risk of infections (25). Here, we focused on describing the taxonomical and functional signature of ESBL-E carriers gut microbiome. Since we did not identify any other covariates influencing the composition of the gut microbiome in this cohort, we may consider this cohort as homogeneous. To our knowledge, this is the first study that intend to explore the specific patterns of the gut microbiome associated with ESBL-E carriage using shotgun metagenomic sequencing, providing a deeper description of bacterial communities and functional pathways than previous works in the field.
Overall, we found that our results are consistent with previous experimental observations and highlight several mechanisms that may play a key role in the expansion of antimicrobial-resistant in the human gut microbiome (Fig. 4).
Human gut microbiome depleted in SCFAs producers may promote ESBL-E colonization persistence.
A large proportion of the bacteria we found depleted in the gut microbiome of ESBL-E carriers, namely Clostridium hylemonae, Collinsella tanakaei, Johnsonella ignava and Bifidobacterium adolescentis have been identified as directly involved in SCFAs production such as acetate, propionate and butyrate, supporting the hypothesis that a depletion in SCFAs producing species promotes ESBL-E colonization and persistence in the gut. (26–28).
Moreover, the only under-represented metabolic feature in ESBL-E carriers was pullulanase (K01200), a glucanase that degrades pullulan, which has been reported to stimulate butyrate production by promoting the growth of Bifidobacteria species. (29). These results confirm in humans the conclusions of a recent experimental work highlighting that SCFAs promote the clearance of EBLS-E through the regulation of intracellular pH (14).
An increased production of succinate by the gut microbiome may promote ESBL-E colonization persistence.
We observed a greater abundance of gene functions involved in glutamate transport (M00233, K10008, K10007 and K10006) in ESBL-E carriers suggesting an increased glutamate metabolism. Glutamate is involved in the biosynthesis of several proteins and mostly metabolized via the GABA shunt pathway. More specifically the production of γ-aminobutyrate through a decarboxylation reaction of glutamate consumes protons which are subsequently removed from the environment. Thus, succinate production from glutamate through Glutamate Decarboxylase has been reported to be implicated in bacterial acid tolerance by facilitating intracellular pH homeostasis (30). These functional observations were consistent with the enrichment of the gut microbiome of ESBL-E carriers in Prevotella and Bacteroides, two known succinate-producing genera suggesting an increased availability of succinate (30,31). Salmonella typhimurium also metabolize microbiota-derived succinate to grow and colonize the gut lumen and the enrichment of the gut microbiota in succinate is involved in Clostridium difficile infection after antibiotic treatment (32,33).
The gut microbiome of ESBL-E carriers is characterized by a decreased SCFAs production and an increased succinate production.
Here, the most significantly over-represented species in ESBL-E carriers was Bifidobacterium adolescentis. B. adolescentis was also found increased in the gut microbiome of travelers who cleared ESBL-E colonization when compared to travelers who did not suggesting it could play a key role in colonization clearance of antibiotic resistant bacteria. In the same way, travelers who did not cleared ESBL-E colonization harbored a significantly higher proportion of Bacteroides sp. in their gut microbiome such as ESBL-E carriers in our study, suggesting a role for this genus in colonization persistence of antibiotic resistant bacteria (34). Altogether, these data suggest that modulations of the gut microbiome leading to an increased succinate production and a reduced SCFAs availability could confer a beneficial environment for ESBL-E colonization.
The gut microbiome of ESBL-E carriers has a greater ability to metabolize multiple microbiota and mucus layer derived carbohydrates.
Our data also show that the gut microbiome of ESBL-E carriers expresses a greater diversity of energy metabolism pathways with an over-expression of gluconeogenesis (M00003), glycolysis (M00001, M00002) or pentose-phosphate pathway (K07404). The increased activity of amino-acid metabolism, especially isoleucine and methionine (M00570, M00019, M00338) and gluconeogenesis (M00003) suggests an intense energy metabolism activity involving alternative pathways that use amino-acids instead of carbohydrates to generate glucose as it has been observed for Escherichia coli in urinary tract infections (35). This increase in the number of energy metabolism pathways in the gut microbiome of ESBL-E carriers suggest a greater ability to metabolize multiple microbiota-derived nutrients, which could be one the mechanisms providing a growth facility to some species such as Enterobacteriaceae and promote the colonization as previously described for Escherichia coli O157:H7 maintenance in the microbiota which involves gluconeogenesis (36). We observed an increased abundance of glycoside hydrolase in the microbiome of ESBL-E carriers. Especially, the over-expression of glycoside hydrolase GH 95 and GH 20 suggests that the microbiome of ESBL-E carriers presents an increased mucin degradation activity resulting in an increased availability of host-derived sugar, such as fucose, which has been shown to facilitate post-antibiotic expansion of enteric pathogens (37,38). Thus, changes in the gut microbiome of ESBL-E carriers supports that colonization is enabled by a favorable nutrient environment.
The gut microbiome of ESBL-E carriers harbors over-represented bacterial mechanisms facilitating bacterial survival and invasion.
Metabolic pathways suspected to be part of bacterial mechanisms involved in colonization were also over-represented in ESBL-E carriers. Among them, we identified an increased abundance of pathways involved in sulfur cycling (M00436, K15553, K15555), supposed to increase bacterial resistance to antibiotics and to protect them from reactive oxygen species (39).
Limitations
Our study has several limitations, the main one being the low prevalence of ESBL-E colonization in our cohort, the reason why we used bootstrap iterations. Second, although subjects included in our cohort were exposed to controlled medication and received the same dietary, the same hygiene and care support, all the parameters that may affect gut microbiome composition, such as visits by outside individuals, were not controlled. Third, this study only provides information on the characteristics of the gut microbiome allowing for persistence of colonization but neither explains the initial changes that occurs nor the mechanisms leading to the acquisition of these resistant strains, as longitudinal design could have done. Also, because this work is based on DNA sequence reads analysis, it only provides informations about metabolic coding capacities. Further investigations, including transcriptomic and metabolomic analysis are needed to confirm the upregulation of these metabolic pathways and the increased availability of succinate in the gut lumen of ESBL-E carriers.