I really try to avoid giving scientific papers the full Mad Biologist treatment, so I was going to let this paper about antimicrobial resistance in space (BOOGA! BOOGA!) lie. And then I saw RT (not a big fan in general of RT) headline a story about the paper with ” Mutant superbugs menace future space station expeditions – NASA.” Not just ordinary superbugs, but MUTANT superbugs (ZOMG11!!11).

Ugh.

So I’ll mostly try to lay off the snark, but I do want to discuss this paper because it highlights some common ways that one should not analyze bacterial genomes to find antimicrobial resistance (AMR) genes. Before I get to the specifics of the paper, it’s worth considering what our expectations should be. It would be very surprising if all space station bacteria lacked AMR genes. Why? Because if you look at human associated bacteria, especially the ‘weedy ones’ (like those characterized in the paper), a subset will have AMR genes. So it really shouldn’t be shocking to find bacteria with AMR genes–I would doubt the validity of a study that didn’t find some bacteria with some AMR genes.

Onto the paper, comrades! In the paper, the authors sequence the genomes of five Enterobacter bugandensis isolates found on the space station. The authors then assembled (reconstructed the genome) and annotated (identified the genes on the genome) these genomes. They then come up with a large list of AMR genes. Before I continue, again, I’m not trying to drag the authors–these are common errors (reviewers and editors, do better).

The first problem is that many bacteria have AMR genes as a matter of course. They might or might not be expressed (‘turned on’), but they are there. In Enterobacter, all isolates, including the five in the paper, have a chromosomal beta-lactamase, blaACT, which can confer resistance to ampicillin and many older cephalosporins. Note the word can; it’s doing a lot of work. Depending on the isolate, the ACT beta-lactamase is or isn’t turned on.

But the key point is that most Enterobacteriaceae, which include E. coli and Enterobacter, have chromosomal beta-lactamases (interestingly, Salmonella lack a chromosomal beta-lactamase*). E. coli have a chromosomal betq-lactamase, blaEC, which, when turned on, works against a large number of beta-lactam antibiotics. Fortunately, it’s usually not turned on. In fact, twenty years ago, most microbiologists incorrectly predicted. that resistance to the newer beta-lactam antibiotics (derivatives of penicillin that start with ceph- or cef-) would evolve through the overexpression of these chromosomal beta-lactamases, and not through the acquisition of novel beta-lactamases (which is what happens much of the time, though less so in Enterobacter). Very long story short: having blaACT isn’t exactly a crisis, it just (usually) means you have Enterobacter (just as blaEC simply means you have E. coli). Get thee to a microbiologist…

The other common problem, and one in this paper, is that point mutations get conflated with genes. What I mean is there are genes in which single mutations can lead to antibiotic resistance. That does not mean, however, that the presence of these genes in and of themselves does not say anything about antibiotic resistance. For example, the paper lists “Translation elongation factor Tu” as an AMR gene. Mutations in EF-Tu can result in resistance to kirromycin, but EF-Tu is an essential protein: without it, the cell dies. Every Enterobacter has this gene. This would be like stating gyrases, which are an integral part of the DNA replication machinery, are AMR genes… oh no, they didn’t… Yes, yes, they did. There also are many other genes (the Mar proteins) that, while not essential, are found in virtually every isolate. Reporting these doesn’t mean much.

Unfortunately, most science reporters (and apparently many reviewers and editors [cough] [cough]) don’t know this, so we start getting stories about SPACE BACTERIA OF DOOM. That’s not to say there isn’t some interesting things: the copper resistance genes seem ‘real’. Whether the copper resistance genes are a defense mechanism against phagocytosis or copper as an antimicrobial is unclear (and these aren’t mutually exclusive), but it’s interesting.

Anyway, when you’re doing these sorts of analyses, if you don’t have the relevant expertise (it’s ok if you don’t!), get thee to a biologist who does. And that goes for reviewers and journal editors too.

*Meanwhile, multiple species of Cronobacter have a chromosomal beta-lactamase that confer resistance only to cephalothin, and no other beta-lactams. No idea how that evolved.