Rough notes from the webinar The Scientist: Precision Medicine Learning Lessons from the Microbiome (now available online)
When studying the human microbiome, it helps to have an “artificial gut” that simulates the microbial environment in a way that allows direct experimentation in vitro. Since the 1990s, people like McFarlane, and later Freeman, have used a “three vessel” structure to isolate and study the microbes of interest. But one of the webinar speakers, Robert A. Britton from Baylor Medical College, found this too restricting and developed a much more scalable verion that lets him study more than 96 of these “guts” at a time. That’s the basis of some of his new discoveries of the nasty pathogen C. dificile.
Originally, C. difficile was thought to be caused by a specific antibiotic, clindamycin, which was found to be linked to these resistant infections. Even in very low quantities (just a few hundred microbes), C. difficile can grow exponentially in the presence of this antibiotic, turning into a serious infection in just a few days.
Closer research reveals that that there are two main “epidemic ribotypes” involved in these infections, including Robotype 027/BI/NAP1 (from the US) and Ribotype 078 (from European livestock)
Dr. Batton and his team, using his artificial gut, noticed that, interestingly, the so-called 027 strains are particularly more fit than others, and the quest to understand the reasons led him ultimately to the sugar trehalose, after experimenting with phenotype microarray plates to understand which specific carbon sources were behind its growth.
Trehalose is a “non-reducing” sugar useful as a food stablizer. You can put more of it into food than sucralose. The body can metabolize it, and is common naturally-ocurring in things with yeast. Trehalose was first launched commercially in 1994, after a new manufacturing method was developed, and approved in 2000 for GRAS – just before the current big spike in C. difficile infections.
Further research showed that every C. difficile has a Trehalose operan, and in all 027 strains (but no other), there is a single L1721 amino acid mutation.
Several people asked questions afterwards. Here are some highlights:
Rob Knight: low diversity is bad, but we only think higher diversity is better.
Which diseases are proven to respond to FMT:
C. difficile (yes); IBD/Crohn’s seems to depend on the donor. Autism looks promising; but most others have been in non-peer-reviewed forums.
One barrier is that the FDA now requires FMT to be registered like a new drug application, which will slow the research.
Interestingly, sometimes FMT works even on sterilized samples, leading to the possibility that maybe the active substance is a metabolite or a phage?
How can you prime yourself to be more resistant to C. difficile? Answer: fermented milk (kefir?), but really there’s no data about one probiotic or another.
Clinical methods for FMT: “northern route” (tube through stomach) and “southern route” (seems to work best)
maybe anaerobic microbes are more necessary for non-C diff conditions, so maybe we can develop a fully-anaerobic technique for FMT
Probiotics: So far, most supplements are based on pre-existing microbes, but in the future it’ll be possible to synthesize new ones, and that will create regulatory issues.
Britton: very hard to tell function based on 16S, even if you can get at the species because some of these are strain-specific.
Don’t use 16S if you want strain-level ID. 16S is okay for ooverall community look, though.
What percent are C. diff in healthy vs unhealthy? “not sure if you can classify somebody as healthy based on the percentage the test shows”
infants seem to be highly colonized but don’t get disease. “we don’t completely know”. don’t know whether it comes from environment or from within you. Even when you’re infected, it doesn’t seem to take over like citrobacter (80%). It’s usually a fraction of a percent even when causing disease.