Many bacteria from deep infections reach the blood through the lymph vessels, a mouse study finds

BESIDES ITS blood vessels, the body has a second network of vessels. It collects the fluid that leaks out among the cells, carries it through a chain of small filtering organs and returns it to the blood. That network is the lymphatic system, and the filters set along it are lymph nodes, the places where immune responses against whatever the fluid carries are organized. Doctors have generally treated the first node that an infected limb drains into as a trap, a place where bacteria are held and killed, so that a serious infection was thought to reach the blood by breaking directly into blood vessels at the wound. A preprint from Imperial College London, updated and posted again on bioRxiv on October 26th, reports that in mice most of the common bacteria tested are not held there. They travel from node to node and into the blood.

Bacteria in the blood can set off sepsis, a reaction throughout the body that can damage organs and kill. The authors estimate that bacteria which live outside the body's cells, the kind studied here, account for about 6 million deaths a year, and they cite published work finding that more than a quarter of bloodstream infections have no identified starting point in the body.

The team, led by Matthew Siggins and Shiranee Sriskandan, injected bacteria into the thigh muscle of female mice and counted the live bacteria that could be recovered a few hours later from the muscle, the blood, the spleen, the liver and several lymph nodes. They used twelve strains taken from patients, spanning six species, along with reference and mutant strains. The species were Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Streptococcus pyogenes and Streptococcus agalactiae, known as group A and group B streptococcus, and Staphylococcus aureus.

The design relies on two nodes near the mouse's foreleg. The axillary and the brachial nodes sit beside each other and receive the same blood supply, but only the axillary one lies on the path that drains the infected hind leg. Bacteria arriving through the blood should reach both nodes in similar numbers. Bacteria traveling along the lymph vessels should reach mainly the draining one.

Six hours after infection, every strain of E. coli, Klebsiella, Pseudomonas and the two streptococci had been recovered from the nodes nearest the infection, from the distant draining node, from the blood and from the organs. The non-draining node held markedly fewer bacteria even in mice whose blood carried large numbers, and for every species the counts were significantly higher in draining than in non-draining nodes. The authors read that as evidence of movement along the lymph vessels rather than seeding from the blood.

Staphylococcus aureus behaved differently. Three strains of it, including the widely used USA300 reference strain and two bloodstream strains from patients, stayed largely in the first nodes and spread little further. That species is the one on which much of the earlier bacterial work rests, including a 2018 study from the University of Calgary showing that immune cells arrive in the draining node within hours and stop it spreading. The authors conclude that S. aureus is "an important exception rather than a representative model".

Some bacteria wear a coat of sugars attached to their surface, called a capsule. A group B streptococcus engineered without its capsule reached the nodes and organs in much smaller numbers than the ordinary strain, while the count at the injection site was no different, which suggests the capsule assists movement rather than survival in the muscle. Digesting the different, secreted sugar layer of Pseudomonas changed nothing.

The caveats are substantial. These were mice, given a large dose of 100m bacteria into muscle rather than acquiring an infection through a wound, in groups of four to six animals and sampled at one or two time points. The authors argue the dose is in the range measured in human pus and wound samples, and that the spread varied widely between species given the identical dose. The work has not yet been peer reviewed. It was first posted in April 2025, and the bioRxiv record does not state what changed in this revision.

If the route turns out to operate in people, it offers an anatomical account of how a limited skin or muscle infection becomes a bloodstream one with no obvious source, and it makes the lymph nodes a place to study when asking how such infections begin.