Premature infants born before 28-30 weeks of life are at high risk for many complications, among which the chances of dying from an infection starting in the intestine are very great. A new study published in the journal Nature Medicine reports that such infections are very much more likely if something interferes with the normal microbial accumulation in the gut. Adding certain bacterial strains protects against such infections, on the other hand.
Researchers Jeffrey Singer and Casey Weaver say, “Our findings offer the possibility for rational design and testing of effective probiotic therapies to protect at-risk premature infants.”
Late-onset infection and dysbiosis
Infections that occur 3 or more days after birth are labelled late-onset sepsis. The researchers concluded that late-onset sepsis is often promoted by the hyper-proliferation of a single species within the gut. This phenomenon is called neonatal dysbiosis. Once this occurs, late-onset sepsis frequently follows. Clinical work showing this finding has now been validated by the findings of this study.
The scientists say, “Our model should help define mechanisms by which pioneer species of the developing microbiome of neonates prevent — or fail to prevent — dysbiosis that predisposes to LOS.”
The current study used mice because newborn mice have intestines which are still developing at birth and shortly afterwards. This makes them more comparable to the immature intestines in very premature babies.
The researchers used newborn pups exposed to a disease-causing strain of the bacterium Klebsiella pneumoniae Kp-43816, which was inserted into their immature stomachs. These microbes have been engineered to carry a biomolecule that is luminescent. As a result, the position of the bacterial growths within the intestine of live animals can be clearly seen. Moreover, it is easy to track their spread out of the gut to invade the rest of the body. Bioluminescence is ideal for this task as it is visible even when overlaid by almost one solid inch of tissue.
Bacteria Klebsiella, 3D illustration Credit: Kateryna Kon / Shutterstock
They inserted the pathogenic strain at a dose that would kill 50% of the animals within 10 days. At this dose, they saw how Klebsiella moved out of the intestine at precisely those points where it had formed the densest growths or colonies. At the same time, dysbiosis was not universally associated with sepsis. This correlates with human premature babies, not all of whom develop late-onset sepsis.
The next step was placing a non-disease-causing strain of the same bacterium, namely, Kp-39, into the newborn mouse stomach. At this point the researchers were astonished to see how this strain did not cause disease, as expected, but were found to move out of the intestine into the abdomen. It also caused infections of the liver or the mesentery, the vascular stalk that connects the abdominal blood vessels to the rest of the body. Over the course of 9 days, the pups cleared the bacteria from their bodies.
The researchers feel that while both Klebsiella strains could migrate through the gut wall, there are clear differences in the way the host immune system clears these microbes from the body after it accomplishes this step. This difference persisted when the strains were injected straight into the peritoneum, so that they did not need to migrate out of the gut lumen.
Dysbiosis with and without late-onset sepsis
In the case of peritoneal injection with Kp-43816, all the newborn mice died within 24 hours. However, when Kp-39 was used, not a single mouse pup died, and the bacteria was cleared from the body within a week. Capsular variations between these strains showed the researchers that the capsule around the more virulent strain resisted the action of immune cells in engulfing and removing these bacteria. As a result, they were able to successfully infect the pus, unlike Kp-39.
As a result of this experiment, the scientists were able to distinguish and track two different events: while the Kp-43816 strain follows a pathway that mirrors late-onset sepsis resulting from dysbiosis in the newborn, the Kp-39 allows the development of dysbiosis without any sepsis or death, which would spoil the picture.
They followed up this study by changing the composition of the gut microbiome in mouse pups. They found that they could alter the level of susceptibility to gut dysbiosis and late-onset sepsis. It is already known that the normal or healthy gut microbiome is crucial in preventing pathogen colonization and invasion. When a population of germ-free mouse pups, which lack an intestinal microbiome, was exposed to Kp-43816, all were alike infected by the bacterium, resulting in late-onset sepsis.
Antibiotic treatment in the mother
Taking things further, the researchers now administered antibiotics to pregnant mice from one day before delivery to several days thereafter, after which Klebsiella infection with the virulent strain was introduced. Two different antibiotics were given, viz, gentamicin and vancomycin. Since both of which are absorbed very little, the presence of antibiotic is unlikely to be significant in the mouse pups. However, the mother’s intestinal bacterial population would be changed in many ways. In turn, this will mean the pups are exposed to a different type of gut microbiome after birth and will develop this type of microbiome.
What happened was that the pups of mothers on gentamicin were very prone to infection, but not those of mothers treated with vancomycin, when compared with controls. Repeating the experiment with the avirulent strain Kp-39 resulted in zero mortality, but greater dysbiosis in gentamicin pups compared to the vancomycin pups.
Lactobacillus strains were more abundant in the intestinal flora of the vancomycin pups compared to the gentamicin pups. Thus, it seems that Lactobacillus numbers correlate with increased or decreased vulnerability to the onset of dysbiosis and late-onset sepsis in the newborn.
DNA sequencing to identify the bacteria that were dominating the intestinal microbiome showed that Lactobacillus murinus was predominant in the pups born to mothers on vancomycin. However, with gentamicin, this strain was almost absent in the pups.
Another intriguing finding was that L. murinus is sensitive to the action of gentamicin but resistant to the action of vancomycin. The removal of this protective and beneficial species by gentamicin administration paved the way for dysbiosis in newborns, by blocking the passage of this ‘good’ bacterium from the mother’s intestine to that of the pups.
Preventing dysbiosis with beneficial bacteria
To confirm these findings, the researchers now administered L. murinus to pups born to gentamicin-treated mothers, before exposing these pups to Kp-39. The incidence of dysbiosis went down sharply in this group compared to controls. In another set of experimental mice, they also found that a strain of Escherichia coli (E. coli) bacteria used in probiotics is also able to confer similar protection in this groups of pups, but the same action was missing when several other common strains of Lactobacillus used in probiotics were tested.
The study also found that very young pups resembled premature babies in that facultative anerobes dominated their intestinal environment. These bacteria belong to different strains but are alike in that they can live and grow with or without oxygen. This is characteristic of the intestines of very immature pups and babies.
With growth and maturation, the microbiome composition changes until it is comparable to that of older or term infants, with predominantly obligate anerobes. These bacteria grow only if oxygen is absent and die in the presence of oxygen. Older pups that had this type of mature microbiome did not develop neonatal dysbiosis.
Thus, it seems that when oxygen is present in the gut of newborn mice, obligate anerobes are inhibited, allowing the overgrowth of other species. This may be why dysbiosis occurs more commonly in very small infants.
The authors conclude that some normal bacterial species are essential as well as adequate to prevent the overgrowth of pathogenic species in the newborn gut, and thus inhibit late-onset sepsis as a result of dysbiosis. Researchers Singer and Casey Weaver say, “They provide a basis for understanding why some probiotics are protective, whereas others are not. This may have important implications for clinical practice, where both maternal and neonatal antibiotic use can alter the neonatal microbiome, and where very-low-birthweight infants are given probiotics without clear evidence as to preferred probiotic species.”
Singer, J.R., Blosser, E.G., Zindl, C.L. et al. Preventing dysbiosis of the neonatal mouse intestinal microbiome protects against late-onset sepsis. Nat Med 25, 1772–1782 (2019) doi:10.1038/s41591-019-0640-y, https://www.nature.com/articles/s41591-019-0640-y