There are a lot of things in human medicine that make sense broadly, but not in detail. We understand why a thing could happen, but not exactly how it happens. A case in point in alcoholic liver disease. It makes perfect sense that longterm alcohol abuse would damage the liver – it’s the front line for everything coming out of the gut, since the hepatic portal vein system drains whatever gets absorbed from the digestive system straight through the organ for first-pass sorting and processing. Continued expose to excess amounts of ethanol cannot help but cause damage, and damage is what you see. But just how does that damage take place?
This new paper (commentary here) from a large multinational team has what may be an unexpected answer: bacterial infection. There have been mouse experiments that suggested that Enterococcus faecalis (a common human-associated bacterial species) might be involved in alcoholic hepatitis, but this latest work provides a great deal of proof that it might in fact be a causative agent. Presence of the bacteria in human stool is strongly correlated with the disease. More particularly, it’s presence of the E. faecalis cytolysin, a bacterial toxin that’s one member of a large family of similar proteins. This particular cytolysin is a mixture of polycyclic peptides, with a number of modified amino acids and lanthionine linkages, and as the name implies, it’s good at causing cell lysis. Unusually, it not only works on eukaryotic cells, but other bacteria as well. This is very likely through the formation of artificial membrane pores in the target cells, but the details are still not well worked out.
One of the more eyebrow-raising takeaways from the new paper is that 89% of the patients with cytolysin in their stool samples died within 180 days of hospitalization, but only 3% of alcoholic hepatitis patients without it died over the same period. This situation can be recapitulated exactly in mouse models, using regular strains of the bacteria versus cytolysin-deficient ones and then adding alcohol to their diet. Even further, using gnotobiotic (germ-free) mice (expensive little beasts, those are) and colonizing them with human stool samples (with and without cytolysin) showed exactly the same pattern.
Down at the cellular level, the cytotoxic effects of the peptide were dose-responsive in pretty much the same way whether or not the cells had been sampled from normal mice or alcohol-fed ones. And that implies that it’s not some sort of cumulative alcohol damage to the hepatocytes that’s causing the problem, but perhaps that the alcohol has an acute effect on membrane integrity and susceptibility to the toxin, and probably to bacterial infection in the first place. The bacteria (and the cytolysin) could be detected in the animals’ liver tissue after deliberate infection in mice dosed with alcohol, but not after deliberate infection with control animals.
Interestingly, the paper goes on to look at bacteriophage therapy. The team isolated several bacteriophage strains that selectively attack E. faecalis, and administered a cocktail of these to infected animals (with a phage preparation against a non-mouse-colonizing bacterium as a control). And indeed, the amount of liver damage in the treated animals was notably decreased. An even more stringent control, phages that target non-cytolysin-producing strains of the bacterium, had no effect.
So this seems like pretty solid evidence that we can add alcoholic hepatitis to the list of diseases that have a bacterial cause, and that suggests immediate possibilities for treatment. In this case, of course, the bacteria are getting a major assist from continued alcohol exposure, and the severity of the disease is correlated with what variety of bacteria are involved. The cytolysin-producing ones are truly bad news, as is the case with bacteria of many other types (as witness cholera, the nastier forms of pneumonia, and more; there are lots of different cytolysins out there).
This work has a lot of intrinsic scientific interest, but I will also confess to a personal connection. As some longtime readers will know, I have unfortunately had a chance to see the effects of longtime alcohol abuse up close (warning, sad post). So I have reason to cheer on anything that can help to deal with the addiction and its consequences!