Introduction

Over the past few years, there has been a growing awareness that the gut microbiota could be an important modulator of pulmonary inflammatory diseases [1–3]. This so called gut-lung axis appears to play an active role in pathogenic processes underlying chronic obstructive pulmonary disease, asthma, acute respiratory distress syndrome and pneumonia [4–9]. Generally, a diverse microbiota has been associated with protection from lung inflammation, whereas low intestinal microbial diversity and antibiotic use are associated with incidence and course of pulmonary disease [10].

Lipopolysaccharide (LPS), a major constituent of the outer membrane of gram-negative bacteria, is recognized by Toll-like receptor (TLR)-4 which plays a key role in initiation of the pulmonary inflammatory response. LPS-induced acute lung injury (ALI) in mice is an often used and suitable preclinical model to investigate pathophysiological features of pulmonary inflammation, applicable to multiple diseases [11, 12]. The response to sterile inflammation after gut microbiota depletion has been studied recently in mice [13]. This study it showed that disruption of the microbiota prior to mechanical ventilation resulted in aggravated inflammatory responses and an increased susceptibility to ventilator induced lung injury (VILI). However, to the best of our knowledge the role of the gut microbiota in the host response during LPS-induced lung inflammation has never been studied. Specifically, we set out to investigate whether disrupted gut microbiota affects the dynamics of key cytokines in lung immunity. The aim of the present study was therefore to investigate the effect of antibiotic induced intestinal microbiota disruption on response to LPS induced acute lung inflammation.

Materials and methods

Results

Antibiotic treated and control mice were given 10 μg LPS from K. pneumoniae, intranasally (Fig 1A). Miseq sequencing and subsequent analysis using a rarefaction depth of 20000 reads confirmed antibiotic induced microbiota disruption depicted by significant loss in diversity (Fig 1B; left panel) and further illustrated by marked different profiles of Family level relative abundances (Fig 1B; right panel). As expected, LPS administration provokes a significant increase in levels of interleukin (IL)-6 but and tumor necrosis factor (TNF)-α (Fig 1C). In BALF, significantly increased levels of interleukin IL-6 but not of TNF-α was observed in antibiotic pre-treated mice compared to controls, at both 2 and 6 hours post induction (Fig 1C). Analysis of cells BALF revealed no significant differences in total numbers as well as in neutrophil population (Table 1). MPO levels in lung homogenate, a reflection of total number of neutrophils in the lung, was significantly decreased in the antibiotic treated group compared to control mice (Table 1). Furthermore, BALF total protein content, a marker of inflammation and vascular leakage, was not significantly different between groups (Table 1). In blood plasma, we found significantly increased levels of IL-6, TNF-α and Monocyte Chemoattractant Protein 1 (MCP-1) at 2 hours after LPS induction, regardless of ABX pre-treatment (S2 Fig). This effect was not observed at 6 hours after LPS induction. No effect of ABX on protein level increase in blood plasma was observed at either timepoints.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Experimental design, microbiota disruption and pulmonary cytokine production during LPS-induced inflammation in control and antibiotic pre-treated mice.

A) Schematic overview of performed experiments. Wildtype C57BL/6 mice were treated with two weeks of antibiotics (ABX: ampicillin, neomycin, metronidazole and vancomycin) in their drinking water after which 1 or 10 μg lipopolysaccharide (LPS) from either K. pneumoniae or E. coli, or saline was administered intranasally. Mice were sacrificed 2 or 6 hours after LPS challenge. B) Gut microbiome diversity as measured by Shannon α-diversity of antibiotic (ABX) treated (yellow) and control (grey) mice (left panel). Group averaged relative abundance profiles on Family level between antibiotic treated and control mice (right panel). Both analysis were executed using a rarefaction depth of 20000 reads (n = 12 No ABX, n = 4 Yes ABX mice. Note: The rest of ABX treated mice (n = 12 total) tested did not reach 20000 reads and were therefore excluded from Fig 1B). C) Interleukin (IL)-6 and tumor necrosis factor (TNF)-α were measured in broncho-alveolar lavage fluid of antibiotic pre-treated mice (yellow) or control (grey) sacrificed 2 or 6 hours after intranasal inoculation with 10 μg LPS from K. pneumoniae. Additional control groups are saline inoculated mice with or without antibiotic pre-treatment (white and black bars, respectively). Dashed line represents detection limit. Data are presented as bar plots showing the mean and standard deviation of the mean (n = 6–8 mice/group). One way ANOVA with post-hoc Tukey’s test. na: not applicable, ns: non-significant; *p<0.05, **p<0.01, ***p<0.001,****P<0.0001.

https://doi.org/10.1371/journal.pone.0241748.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Cellular influx and protein content of bronchoalveolar lavage fluid during LPS-induced inflammation at t = 6 hours.

https://doi.org/10.1371/journal.pone.0241748.t001

We repeated these experiments using 10 μg LPS from E. coli. A similar trend of increased TNF-α and significantly increased IL-6 levels at t = 6 in BALF was observed (S1A and S1B Fig) Again, pulmonary neutrophil recruitment was not different in control and antibiotic pre-treated mice (S1 Table). Finally, in order to evaluate dose dependency, we repeated this experiment with 1 μg of LPS. This showed similar results (S1 Fig), with the notion that the increase of TNF-α levels at t = 6 in BALF derived from microbiota disrupted mice reached statistical significance in this experiment (S1B Fig).

Discussion

To our knowledge, this study is the first to evaluate the effect of antibiotic induced disruption of the intestinal microbiota on LPS-induced lung inflammation. We observed a modest but consistent effect of antibiotic pre-treatment on pulmonary cytokine production after a challenge with LPS from different gram-negative bacteria. No differences in cytokine levels in the systemic compartment were observed. Overall our data suggest that microbiota disruption might predispose to exaggerated inflammatory responses, which was reported before in several mouse models [4, 5, 8, 13]. A healthy, well balanced microbiota might therefore be important for modulating appropriate inflammatory responses. The exact underlying mechanisms remain to be elucidated, however, it is suggested that the gut microbiota is important in the development of healthy innate immunity [19]. Furthermore, antibiotic disruption was found to impair murine haematopoiesis and deteriorate outcome in a pneumonia model [4, 20]. IL-6 in the lungs is produced by both immune and nonimmune cells after various stimuli and can have a wide range of effects. Increased levels of this cytokine have been implicated in various (pulmonary) inflammatory diseases. For instance, IL-6 is increased in BALF and sputum of asthma patients. Besides being present as an immune modulator in various lung diseases, IL-6 has also been suggested to play a role in pathogenesis of asthma and COPD [21]. The observation of decreased levels of MPO in homogenates of antibiotic treated animals lung might indicate immature neutrophils as a result of impaired hematopoiesis. This hypothesis, however, should be addressed in more detail in follow-up studies. Overall, these results are in line with previous reports on altered pulmonary immunity after interventions affecting the intestinal microbiota, in models of pneumonia, allergic airway disease and mechanical ventilation [4, 5, 8, 13].

This study has a number of limitations. Although our antibiotic regimen with known effect on gut microbiota included non-absorbable antibiotics [4], we cannot completely rule out a systemic, potentially immunomodulatory, effect. In the current study, however, we found no significant differences on systemic parameters between antibiotic treated and control mice. Furthermore, we have previously assessed this regimen’s effect on lung microbiota composition, and found no differences between antibiotic treated and control mice [3]. In addition, mice from a different vendor may have a different microbiota and might therefore display different inflammatory responses, as was demonstrated for sepsis as well as malaria severity [22, 23].

In conclusion, antibiotic induced microbiota disruption is associated with alterations in host responses during LPS-induced lung inflammation. An increased pro-inflammatory, IL-6 driven response after treatment with antibiotics could thus contribute to development or severity of pulmonary diseases. Further studies, addressing the effect of gut microbiota depletion on immune cell function, are required to determine the underlying mechanisms as well as the potential clinical relevance of the gut-lung axis in pulmonary infection and inflammation.

Supporting information

Acknowledgments

The authors would like to thank Marieke ten Brink and Joost Daalhuisen (Academic Medical Center, Amsterdam, The Netherlands) for their indispensable technical assistance during the mice experiments.