Early antibiotic use disrupts immune development in infants

A new study by researchers at the University of Rochester Medical Center (URMC) has found that early exposure to antibiotics may impair the development of an infant's immune system, and a natural metabolite may be key to reversing the damage.

A study published in Cell found that exposure to antibiotics during pregnancy and infancy can permanently weaken the immune system’s ability to fight respiratory infections like the flu. Analyzing both mouse models and lung tissue from human infants, the researchers found that early antibiotic use disrupts the gut microbiome’s ability to produce inosine, a molecule that serves as an important signal for developing immune cells.

However, by adding inosine to mice, the scientists were able to correct the immune system problems caused by antibiotics, opening the door to potential therapeutic strategies to boost immune memory in vulnerable infants.

“Think of inosine as a molecular messenger. It travels from the gut to developing immune cells, ‘teaching’ them how to properly mature and prepare for future infections,” explained Hitesh Deshmukh, MD, PhD, senior author of the study and chief of neonatology at Golisano Children’s Hospital (GCH) at UR Medicine.

The project was part of a long-term NIH R35-funded initiative to study how early-life exposures shape lifelong risk for diseases including asthma and chronic lung disease.

“We know that antibiotics can save babies’ lives, but they also disrupt the microbiome during a critical period of immune system development,” Deshmukh said. “Our study shows one way this disruption affects lung immunity and, more importantly, a possible way to correct it.”

The disorder affects the formation of tissue-resident memory T cells, a specialized population of immune cells that live in the lungs and provide long-term protection against viral infections. Without these cells, infants may remain vulnerable to severe respiratory illnesses well into adulthood.

“We found that the gut microbiome acts as a teacher for the developing immune system,” Deshmukh explained. “When antibiotics disrupt this natural educational process, it’s like removing key chapters from a textbook: the immune system never learns important lessons about fighting respiratory infections.”

Key findings of the study:

The study compared infant mice exposed to common antibiotics (ampicillin, gentamicin, and vancomycin—the same drugs often used in pregnant women and newborns) with mice whose natural microbiome remained intact.

In mice exposed to antibiotics:

• The population of protective CD8+ T cells in the lungs was significantly reduced.

• There was an impairment in the ability to form tissue-resident memory cells, specialized immune cells that live in the lungs and provide rapid protection against re-infection.

• Immune deficiencies persisted into adulthood, indicating persistent changes in immune system development.

Using lung tissue samples from the NIH-funded BRINDL Biobank, the team confirmed that similar immune deficiencies were present in human infants exposed to antibiotics. Not only did these infants have fewer memory cells, but they also showed gene expression patterns similar to older adults, which is also associated with an increased risk of respiratory infections.

Most importantly, adding inosine to antibiotic-exposed mice significantly restored their ability to develop functional memory cells and mount effective immune responses, opening up promising prospects for future therapies.

“This suggests that we can protect at-risk infants with targeted supplementation,” Deshmukh said. “While much more research is needed before this approach can be applied clinically, we now have a path forward.”

The study's findings could influence future research into developing interventions — including dietary supplements, metabolite therapy, or microbiome support strategies — to help newborns develop stronger immune memory without having to rely solely on antibiotics or risky probiotics.

Deshmukh noted that GCH neonatologist Gloria Preihuber, MD, played a key role in the study. Her BRINDL biobank of NIH-supported infant lung samples collected over a 15-year period allowed the team to test their findings in human cells.

“This paper would not have been possible without Dr. Prayhuber’s generosity and expertise,” Deshmukh said. “Being able to compare the mouse results with human cells was absolutely critical. That’s one of the main reasons I came to Rochester (from Cincinnati Children’s) — to collaborate with her.”