Today I’d like to talk about something closer to home… the human microbiome! I mentioned this briefly in a previous post where I introduced metagenomics, describing how the human microbiome isn’t just a collection of cells that happen to live with us, but a complex community of bacteria, archaea, fungi, and viruses, all coexisting in small, specialized habitats throughout the body. Every person carries around an entire ecosystem of trillions of microbes that live in our mouths, guts, on our skin, and in places we rarely think about.
Using metagenomics, we’ve learned just how diverse and interconnected these microbial communities are, and how disruptions in this balance are linked to certain diseases ranging from metabolic disorders to neurological conditions. It really is a dynamic system, constantly shifting and adapting as different microbes grow, compete, and cooperate in ways that affect our health, metabolism, and immunity, among others. I know I’ve mostly talked about lateral gene transfer (LGT) in environmental contexts, but it’s just as fascinating, if not more, to see it play out within the human body. You might wonder whether the microbes living inside us actually share genes the way environmental ones do. Well, yes, they do.
A recent paper published this year in Nature Microbiology explored this idea in remarkable depth. This study was led by an excellent team of researchers including my PhD supervisor, Rob Beiko, an LGT wizard himself, who helped develop a computational framework called WAAFLE. And no, it’s not the kind you eat for breakfast. WAAFLE actually stands for ‘Workflow to Annotate Assemblies and Find LGT Events’, and is a powerful tool built to detect gene transfers directly from metagenomic assemblies, which is important because traditional approaches rely on analyzing complete, isolated genomes. Metagenomic data are much messier and often fragmented, remember, we’re sequencing DNA directly from an environment, in this case, the human body, where countless species coexist and interact.
WAAFLE was designed to handle that complexity though. It scans through DNA fragments from microbiome samples and spots when a gene seems out of place (remember when we talked about this?), like it came from a different species than the rest of its neighbors. If most of the genes on a DNA fragment belong to one species, but one gene clearly matches another species, then that’s a potential signal of lateral gene transfer. It does this by first predicting genes, then matching them to known species from reference databases. When it finds a breakpoint where the species identity changes along the sequence, it flags that as a potential transfer. WAAFLE can also estimate how often gene transfers happen across different body sites. It’s important to note that this is a big deal, because only a few methods have been developed to reliably identify and profile LGT in microbial communities.
To make sure those signals are real and not just artifacts, WAAFLE applies several filters to check that the classifications are well supported and make sense biologically. It was also benchmarked using simulated datasets where the true transfers were already known, and WAAFLE recovered them with high precision and a false-positive rate below 1.5%, meaning that fewer than 1.5% of the gene transfers it detected were incorrect. This is very good for such complex data!
When applied to more than 2000 assembled metagenomes from 16 body sites, including mouth, gut, skin, and vaginal microbiomes, they identified more than 100,000 high-confidence gene transfer events. The rate of transfer wasn’t uniform across the body; it was highest in the oral microbiome, especially in places like the tongue, gums, and dental plaque, where close contact between cells makes gene transfers easier. The gut followed, while the skin had the fewest transfers, probably reflecting its more transient and exposed environment. They also found that these gene transfers weren’t random. They were more likely to happen between microbes that are closely related and between those that are more abundant in the community, meaning that, who shares genes with whom depends on both how similar they are and how many of them there are. Gene sharing also formed structured networks rather than random exchanges, since most microbes took part in only a few transfers, but a few acted as central hubs that connected many others.
Many of the transferred genes were linked to mobile genetic elements, which I’ve talked about before as the vehicles that move bits of DNA within and between cells. One of the most important findings was how often antibiotic resistance genes appeared among the transferred DNA. These were especially common in the mouth and gut and included resistance to several major antibiotic types. This shows how microbes living within us constantly exchange tools that help them adapt, survive, and sometimes resist treatment. To confirm the accuracy of their findings, the authors additionally tested several predicted transfers in stool samples and found that many were real, showing that these transfers are truly happening within our bodies. This is a reminder that evolution isn’t something that happened long ago and it’s still unfolding within us, in the smallest and most unexpected ways, but we don’t often stop to think about that, do we?
Image by Alexander Fox | PlaNet Fox from Pixabay
