NEW YORK (GenomeWeb) – Using a shotgun metagenomic sequencing approach, a team from Quebec's Laval University has uncovered new data on how the gut microbiome is reshaped by exposure to antibiotics, specifically the drug cefprozil.
In the study published earlier this month in The ISME Journal, the researchers found that while there was great heterogeneity both in the initial states and the responses of participants' gut microbiomes to antibiotic treatment, there were also some repeated patterns that emerged, such as an association between initial low diversity in the gut microbiome and post-antibiotic increases in potentially harmful bacteria that could have implications for personalizing antibiotic treatment.
Attention to the effect of antibiotics or other therapies on the human microbiome is growing, with recent studies finding, for example, that the use of proton pump inhibitors results in lower diversity in the gut microbiome, potentially explaining the known contribution of PPIs to the risk of Clostridium difficile infection.
"Several other studies have been made with antibiotics to see the impact on the microbiome," Frédéric Raymond, the study's first author told GenomeWeb this week.
"But most did not include many healthy people. They either looked at people who were sick who had antibiotics in the hospital, or just had smaller groups of participants. [Either way,] it was not enough to see how antibiotic treatment effects on the gut microbiome are reproducible," he said.
In their own antibiotic study, Raymond and his Laval colleagues took a step further than many others, by attempting to control for as many factors as possible that could influence the fluctuation in normal microbial diversity from person to person.
The researchers applied an extremely stringent selection criteria to recruit a total of 18 healthy volunteers and another six healthy control subjects for sequencing. Participants were excluded if they had been recently ill or were chronically ill, or if they'd taken antibiotics over the past year or a variety of other drugs.
But exclusion criteria also included a variety of other conditions that might affect the baseline gut microbiome, including working in a healthcare facility, farm, slaughterhouse, or animal care facility.
Vegetarians were also ineligible, as were smokers, chronic drinkers, pregnant and lactating women, and those with a high or low body mass index. Even someone simply living with a human or animal that had taken antibiotics in the last month or who fit some of these other criteria was excluded from the study.
After collecting their cohort, Raymond and his colleagues sequenced stool samples from three time points. The 18 primary subjects had samples taken at baseline, then after taking the antibiotic cefprozil for seven days, and then once more 90 days after the end of their course of treatment.
The six control participants were unexposed to any antibiotics but still sequenced at the same three time points.
Adopting a shotgun metagenomics approach — as opposed to targeted 16S rRNA sequencing, which Raymond said the group also considered — the researchers were able to not only look at the presence and concentration of various bacterial populations, but to also compare changes in antibiotic resistance-associated genes pre- and post-treatment.
"We had a lot of internal discussion to see if we would go with the classic 16S sequencing or shotgun sequencing, [and] I worked hard to have everyone on the same page for the metagenomics approach, because one of the goals was to see the impact on resistance genes themselves and to do that we had to sequence everything with a good depth," Raymond said.
Amongst the group's findings was an observation that inter-individual variability in resistance genes was even greater than the variability in bacterial flora, with most resistance genes observed only in a limited number of participants.
According to the authors, this underscores the difficulty in generating cohorts with homogenous microbiomes, even under criteria as stringent as those used by the Laval group.
Despite this observed heterogeneity, the investigators did find that antibiotic exposure had a reproducible effect on participants' microbiomes. Overall, antibiotic use appeared to cause a significant decrease in the level of six bacterial families. In contrast, only three genera systematically increased during antibiotic treatment.
The dominant taxa observed prior to treatment did not appear to be significantly affected by antibiotic use, the authors also wrote. Rather, treatment with cefprozil mostly altered levels of bacteria that were of low abundance to begin with.
The most consistent affect in the cohort, the team found, was an increase of the Lachnoclostridium bolteae — in 16 of the 18 treated subjects. According to the authors, resistance genes related to the species were also apparent in the sequencing data.
L. Bolteae is known to be resistant to the second-generation cephalosporin cefoxitin and has been associated with dysbiotic states associated with a number of diseases or disorders, they wrote.
The researchers also found another interesting pattern amongst a subset of their cohort who showed a large increase in the opportunistic pathogen Enterobacter Cloacae after treatment. Most (five out of six) of these subjects also had a low-diversity Bacteroides enterotype before antibiotic treatment, despite otherwise meeting all the healthiness criteria for inclusion in the study.
"Not everyone that had a low-diversity Bacteroides enterotype ended up with the bloom," Raymond said. "But almost all those that had the bloom initially had the low diversity Bacteroides."
The link between low initial microbiome diversity and post-antibiotic bloom of E. cloacae appears reminiscent of similar patterns with C. difficile infection, he added, and may have implications for treatment decision making.
Moving forward, it will really be important, Raymond said, to understand how many different antibiotics affect the microbiome.
He and his colleagues hope to look at other drugs as well, but are so-far still in the grant-writing process for such studies.