Prior oral non-absorbable antibiotic-induced gut dysbiosis worsens outcomes ofP. aeruginosalung infection.
First, we assessed if prior oral non-absorbable antibiotic modified outcomes of P. aeruginosa acute pneumonia. Antibiotics administered in drinking water seven days before P. aeruginosa intra-nasal challenge were associated with worse infection outcomes (Fig. 1) as shown by increased lung bacteria load (Fig. 1A), increased extrapulmonary bacterial dissemination to the spleen (Fig. 1B) and increased lung injury (Fig. 1C) when compared to untreated controls. Finally, the survival of antibiotic-treated mice decreased compared to untreated controls (Fig. 1D). Overall, acute sublethal P. aeruginosa pneumonia became lethal in antibiotic-treated mice compared to untreated controls.
Second, we assessed if oral non-absorbable antibiotics induced gut dysbiosis. The gut microbiota from antibiotic treated-mice before P. aeruginosa intra-nasal challenge was altered compared to control (Fig. 1E). We also observed an eradication of several bacterial families like Muribaculaceae, Prevotellaceae, and Lachnospiraceae and an expansion of Burkholderiaceae, Clostridiales, and Lactobacillaceae. Likewise, gut microbiota diversity (Fig. 1F) by one log of 16srDNA compared to untreated mice (supplemental Fig. 3). As expected, we did not observe any modification of the effect of prior oral non-absorbable on the composition and the diversity of the lung microbiota before P. aeruginosa infection (Fig. 1G).
Finally, to demonstrate the causality of gut dysbiosis in worse outcomes of P. aeruginosa infected mice, we treated dysbiosis using fecal microbiota transplant (FMT). Following dysbiosis-inducing antibiotics, FMT before P. aeruginosa inoculation restored pneumonia outcomes to those of control mice (Fig. 1A, B, C, D). FMT also restored the composition and the diversity of the microbiota equivalent to controls (Fig. 1E, F) and did not modify the lung microbiota (Fig. 1H). Of note, P. aeruginosa intra-nasal challenge did not further modify the gut microbiota shift observed in antibiotic-treated mice (supplemental Fig. 4). As expected, P. aeruginosa intra-nasal challenge resulted in a lung microbiota dominated by Pseudomonadaceae (supplemental Fig. 4).
Oral non-absorbable antibiotic-induced gut dysbiosis results in widespread lung and spleen immune depression associated with altered myelopoiesis.
Next, we assessed whether antibiotic-induced gut dysbiosis had any effects on baseline lung and circulating cellular immune profiles as a potential mechanism underlying worse outcomes following P. aeruginosa lung infection.
Cell population analysis in lung tissue showed that antibiotics resulted in widespread depression of lung cellular immunity with a significant decrease in alveolar macrophages, cDC2, patrolling monocytes, neutrophils, γδ-T cells, NKs, and iNKT cells (Fig. 2A). FMT restored most of these alterations (Fig. 2A), establishing a significant role of antibiotic-induced gut dysbiosis in this widespread depression of the lung immune response.
Likewise, antibiotics resulted in widespread depression of most studied spleen immune cell populations (p < 0.05 except for a trend in NK cells), also partially restored by FMT (Fig. 2B).
Because the spleen is a hematopoietic organ in mice, we sought to determine the effects of antibiotics on immune cell hematopoietic factors, specifically on circulating serum levels of GM-CSF, M-CSF, and Flt3-Ligand (Fig. 3A). While circulating levels of GM-CSF were at the detection threshold and remained unaltered, antibiotics induced a significant decrease in Flt3-Ligand, and only a trend in M-CSF decrease. Given the known role of Flt3-Ligand as a major hematopoietic stimulating factor, mainly for monocyte and DC progenitors [9], we studied the effects of antibiotics and FMT on bone marrow monocyte and DC progenitors (Fig. 3B). We found that antibiotics were associated with a significant decrease in bone marrow progenitors specific to resident monocytes (MonoLy6C-) and several specific to DCs (total pre-DCs, pre-DCs 1 and 2 and cCD2 biased pre-DCs). Among these, FMT following antibiotics restored levels of pre-DCs 1. FMT also stimulated the expansion of bone marrow progenitors common to monocytes/DCs (MDP), common dendritic cell progenitors (CDP), and specific to monocytes (cMoPs). Our results suggest that the effects of antibiotic-induced dysbiosis on monocyte and DC progenitors may be involved in the widespread antibiotic-induced lung immune depression associated with worse P. aeruginosa lung infection outcomes.
Hematopoietic cytokine Flt3-L stimulates bone marrow progenitor and lung immune cell expansion and restores outcomes ofP. aeruginosalung infection following oral non-absorbable antibiotics.
Systemic Flt3-L following antibiotics (Fig. 4A) stimulated the expansion of several progenitors (MDPs, CDPs, and cDC-1 biased pre-DCs). Flt3-Ligand administration partially restored or overstimulated the expansion of depressed alveolar macrophages, cDC2, patrolling monocytes, neutrophils, NK, and iNKT (Fig. 4B).
Finally, the effects of Flt3-Ligand administration following antibiotics on the outcomes of sublethal P. aeruginosa lung infection were similar to those of FMT: outcomes were restored to levels not significantly different from controls without antibiotics (Fig. 4C, 4D, 4E). Flt3-Ligand was also associated with decreased mortality in a lethal P. aeruginosa lung infection model similar to the effects of FMT (Fig. 4F).